CC BY
14
S •10
1
G.S
G.6
G.4 0.2
G
0.02 0.04 0.06 0.08 0.1
(h)
Рис. 4.11
п — 2
106кг
см2
-; ^20 — 1.6'
106кг
см
0
— 1.8
106кг h. h
--; -i — 0.4 ; -— 0.5;
см h2 b
ft — 1000 • кг/см2•град; #
2
= 800 кг/см2•град; 70 = 300К; Г11 = 300К; 7\2 = 500К;
ÛJ —
Хп
12т
— 0.001
Литература
1. Вольмир А.С. Устойчивость деформируемых систем. м.Наука, 1967, 984 с.
2. Ломакин В.А. Теория упругости неоднородных тел. м.Изд-во МГУ, 1976, 320 с.
3. Шаповалов Л.А. Влияние неравномерного нагрева на устойчивость сжатого стержня. ПММ, 1957, т. ХХ11, в.1., с. 119-123.
4. Зубчанинов В.Г. Об упругопластической устойчивости слоистых стрежней. Прикладная Механика, 1970, в.6, №2, с.127-129.
5. Yang Y.B., Lin T.J., Len L.I. Thermal effect on the Postbuckling Behavior of an elastic or elasto-plastic truss. Journal of Mechanics, vol.134, №4, 2008, p. 330-338.
6. Amin Heydarpour and Mark Andrew Bradford. Nonlinear Analysis of Composite Beams with Partial Interaction in steel Frame Structures at Elevated Temperature. Journal of Structural Engineering. V. 136, 2010, p. 968-978.
7. Voshoughi A.R., Malekzadeh P., Banan Mo.R. Thermal postbuckling of laminated composite skev plates with temperature-dependent properties. J.Thin Walled structuresv.47.N-7 .2011.p.804-811.
8. Thug P.Vo, Huu-Tai Thai. Vibration and buckling of composite beams using refined shear deformation theory.International Journal of Mechanical Sci-ences.V.62.N1.2012.p.67-76.
ON SPECTRAL PROPERTIES OF BI-DOPED SILICA OXIDE GLASS
SYSTEM.
Shulman I.L., Sadovnikova Y.E.
MIREA - Russian Technological University
Abstract. A new Bi-doped Mg-Al-silicate glass is suggested and investigated. The characteristic relaxation time of 300-800 |is in combination with the high quantum yield and wide excitation spectrum makes this glass a promising laser material. The obtained quadratic dependence of visible absorption intensity is an argument in favor of the hypothesis that the absorption and IR luminescence in Bi doped glasses are caused by Bi2 dimers.
Introduction
Since 2001 [1] bismuth and aluminum co-doped SiO2 glass has been known to exhibit under optical excitation an unusually ultrabroadband (ranging from 1.0 to 1.6 microns) emission. The interest in this material has been greatly increased when the first laser action has been demonstrated in Bi and Al co-doped SiO2 fibers fabricated by MCVD technology [2]. Unfortunately MCVD method is not suitable for bulk glass production.
Since 2001 similar ultrabroad luminescence has been discovered in quite various types of bismuth doped (phosphate, borate, germanate, chalcohalide) glasses [3-6], most of them containing aluminum. There was a number of communications about high optical on/off gain in bulk glass samples, but no lasing effect has been obtained in bulk glass till now.
Despite many investigations, the nature of the luminescent centers is still a question. The role of aluminum in formation of these luminescent centers is important but also not clear. For example, [1] attributes the IR luminescence to Bi5+ and [3, 4] - to Bi+ ions. But the opinion about Bi5+ ion contradicts with the fact that the luminescent centers are easily formed in reducing conditions. To the contrary, the addition of a strong oxidizer (CeO2) in the glass prevents the formation of the luminescent centers [7]. Alternative opinions connect the IR luminescence with Bi2+ ions or some ion clusters. For instance, papers [8,9] attribute the visible range absorption and IR luminescence to Bi2, Bi2- or Bi22- ion dimers.
The goal of this paper is to make a contribution to the above mentioned problems by:
- developing a glass composition with substantially lowered (in comparison with silica glass) melting temperature (i.e. suitable for bulk synthesis) combined with spectral properties close to that of MCVD -produced Bi and Al co-doped SiO2 fibers;
- presenting some new experimental data that can shed some light on the origin and nature of the emitting centers.
Glass composition choice, sample synthesis and preparation.
As it was already mentioned, in principle the IR emission can be observed in a very large variety of Bi-doped glass compositions. But not all of them are promising laser materials. For example, now we do not consider the aluminophosphate glass investigated in our
previous paper [7] as a good candidate for the bismuth laser host. The reasons are both technological (unstable replication of its spectral properties because of high volatility of its components) and spectral (that would be considered below).
SiO2 glass (even with a few % Al2O3 doping) requires too high (>1700°C) melting temperature. For this reason it can't be fabricated with high optical quality in bulk by traditional crucible melting technology.
In the present study we deal with aluminosilicate glass with the basic composition 22.2%mol. MgO + 22.2%Al2O3 + 55.6%SiO2 + additional X% Bi2Os doping (X ranging from 0.25% to 2%). This composition can be synthesized at 1550 0C.
It is known that the luminescent properties of Bi-doped glass are highly sensitive to the synthesis conditions (temperature and reducing/oxidizing atmosphere). To ensure the identical thermal history of the samples series with varying Bi concentration they were prepared in a single run of the oven.
The samples (~50g each) were sintered starting from reagent-grade oxide powders in alumina crucibles in air. The synthesis at 1550 +1°C lasted for 4 h. After the synthesis the crucibles were extracted from the oven and cooled in air. The resulting scatter- and inclusions-free pink-colored glass samples were annealed starting from 700°C. 2-mm thick polished plates were fabricated of the synthesized glasses. The optical quality of the plates was moderate because of striae inevitably present in non-stirred glass.
Experiments.
1. Absorption spectra
Fig. 1 presents the transmission spectra of the Bi-doped Mg-Al-Si glass series in comparison with an example of Bi-doped aluminophosphate glass spectrum. The Mg-Al-Si glass samples demonstrate absorption peaks at 500 and 700 nm and a shoulder at 800 nm (same as SiO2 glass doped by Al2O3 and Bi2O3 [1]). The aluminophosphate glass has different spectrum. It has an intensive absorption peak at 450 nm and a peak at 700 nm. The difference in the absorption spectra indicates that the types of color centers can vary in different glasses. As it was shown in [10] the 450 nm absorption peak in aluminophosphate glass is not associated with IR luminescence.
a
c/o
s
0.25% —U=— —
Ir \ / / W /
1.0% /
! , i / 2.0%
400
600 800 Wavelength, nm
1000
Fig. 1. Transmission of the Mg-Al-Si glass sample series (Bi2O3 concentration indicated) in comparison with that
of aluminophosphate glass sample (dashed line).
The analysis of the absorption spectrum in the Bi concentration series of Mg-Al-Si glasses has shown that:
1. The Beer-Lambert law is not valid in this case. As it is shown in Fig.2, the optical density at the peak wavelength 500 nm is proportional to the square of Bi content. An important conclusion can be derived of this
fact. Quadratic concentration dependence is an argument in favor of the above-mentioned hypothesis that attributes the visible range absorption and IR luminescence to optical centers associated with Bi2 dimers. Their equilibrium concentration should be proportional to the square of Bi ions content in the glass.
0.3 0.4 0.5 0.6 0.70.80.9 1 Bi,0, concentration, mol.%
Fig.2. Extinction coefficients of Mg-Al-Si glass at the peak wavelengths 500 nm versus Bi2O3 concentration.
2. Each curve in the logarithmic plot of optical density versus wavenumber (Fig 3) can be brought into coincidence with another one by vertical shift (within the measurements accuracy). In other words, the shape of the Bi-associated spectra in the visible range (400600 nm) is concentration-independent. Stability of the spectra shape means that:
- practically all the visible range absorption is caused by Bi2 dimers;
- the ratio between different types of centers associated with Bi2 (if any) remains
constant in the concentration series of the samples synthesized at the same temperature.
53 e
<l> e o
.2 e":
* e-HJ
2.0%
1.0%
( 5%
0.25%\
12000
14000
16000 Wavenumber, cm"
Fig.3. Extinction of Mg-Al-Si glasses versus wavenumber (Bi2O3 concentration indicated).
18000 20000 ■1
22000 24000
Unfortunately the poor reproducibility of the phosphate glass spectral properties hampered the similar analysis of its concentration series.
2. Emission spectral properties.
It is well known that the emission spectra of Bi doped glasses are dependent on the excitation wavelength. Fig. 4 presents three examples of the Mg-Al-Si glass emission spectra under excitation into the different absorption peaks.
3
ra
J o 5= C O
E
ULl
0.8
0.6
0.4
0.2
0.0
r\
\ \ \ \
\ ^ \ 800 nm
\ 500 nm
\
J 700 nmV
-f
1000 1200 1400
Emission wavelength, nrn
1600
Fig.4. Emission spectra of Bi2O3 doped Mg-Al-Si glass under various excitation wavelengths (indicated).
In order to make an overall view of the emission properties of the investigated glasses and to distinguish the types of emitting centers in them we have carried out 3D plots of the luminescence photon flow versus incident and emitted wavelengths. In these experiments we have used a tunable OPO (SOLAR laser systems
LP603) and calibrated emission registration through the grating monochromator. The plot (see Fig. 5) shows that in the Bi doped Mg-Al-Si glass there are two excitation peaks at 500 and 700 nm causing 1000-1300 nm luminescence and a third peak at 800 nm causing 1200^1600 nm luminescence.
Fig. 5. 3D plot of the luminescence photon flow versus incident and emitted wavelengths for Mg-Al-Si glass with
1%mol Bi2O3.
The similar plot for the aluminophospate glass sample has the first two peaks but no third one. Thus it can be concluded that there are at least 2 types of the emitting centers in the Mg-Al-Si glass.
An important parameter of a laser medium is the absolute value of luminescence quantum yield q. Its measurements for the IR emission in the Bi-doped glasses were carried out in the following way.
Bi203 concentration, mol.%
Fig. 6. Emission quantum yield of Mg-Al-Si glass versus Bi2O3 doping level. Excitation wavelengths 520 and 700
nm (indicated).
It is known that the response of a Ge photodiode in 950^1550 nm spectral range is accurately proportional to the number of incident photons. Thus q is proportional to the time integral of the emission signal detected by such a diode and reciprocal to the number of absorbed pump photons. In order to estimate the absolute value of the quantum yield of Bi-doped glass the signal was calibrated using an etalon sample of Er doped well dehydrated phosphate glass (Emission ~1.5 |im, q=85+5% [11]) pumped at 520 nm. Fig. 6 shows the results of the q measurements for the Mg-Al-Si glass Bi concentration series excited at 520 and 700 nm. It can be concluded that:
1. The quantum yield is different for different excitation wavelengths. Pumping into ~700 nm band is noticeably more efficient than pumping into ~500 nm band. At present moment we have no explanation to
this fact. The absolute values of the emission quantum yield in Bi-doped Mg-Al-Si glass can reach 65+5% under 520 nm excitation and 85+5% under 700 nm excitation. In contrast to Mg-Al-Si glasses the Bi doped alu-minophosphate glass exhibits much less quantum yield (q=5^6% under 520 nm excitation). We do not know any literature data about the concentration dependence of the quantum yield of bismuth IR emission in glasses. The rather high values obtained in this experiment make Bi-doped Mg-Al-Si glass a promising material for bulk and fiber solid-state tunable lasers.
2. The quantum yield q in Bi doped Mg-Al-Si glass decreases with the doping level. The mechanism of this concentration quenching is not investigated still. Nevertheless the presented data can be useful for reasonable choice of bismuth concentration in the laser glass.
3. Emission decay kinetics.
The emission decay kinetics were measured in a wide (1000^1800 nm) spectral range covering all the IR emission bands. We used selective excitations by OPO and two detector types: a slow (time response ~7
|is) Ge photodiode and a fast (time response ~10 ns) Ge avalanche photodiode. The results are presented in Figs 7-9. It should be noted that the decay curves consist of two stages: "fast" and "slow". Both components are always present in this or that ratio at any pump and detection wavelength.
_j e"
1 / 2
4 Sii "-Î
1
0 200 400 600 800 1000 1200 1400 1600 1800
Time, (is
Figs. 7. Emission decay kinetics of Mg-Al-Si glass series. Time response 7 ¿us. Excitation wavelengths: 520 nm. 1 - 0.25mol.% Bi2O3; 2 - 0.5mol.% Bi2Os; 3 - 1.0mol.% Bi2Os; 4 - 2.0mol.% Bi2Os
Figs.8. Emission decay kinetics of Mg-Al-Si glass series. Time response 7 jus. Excitation wavelengths: 700 nm. 1 - 0.25mol.% Bi2O3; 2 - 0.5mol.% Bi2O3; 3 - l.Omol.% Bi2O3; 4 - 2.0mol.% Bi2O3
Figs. 7-8 show the decay curves of the Bi doped Mg-Al-Si glasses detected by the "slow" photodiode under 520 and 700 nm excitation respectively. It can be concluded that:
- the decay curves are slightly different for 520 and 700 nm excitation;
- the "slow" non-exponential decay has characteristic time 300^800 js. The concentration quenching
brought out in the quantum yield measurements correlates with the lifetime reduction.
Fig. 9 displays some examples of the initial ("fast") decay stage detected by the fast photodiode. In all the cases the fast component was nearly exponential with characteristic time of about 2|is.
Fig. 9. Initial stages of emission decay kinetics in Bi-dopedMg-Al-Si glass. Time response 10 ns. Excitation
wavelengths: 520 nm and 650 nm (indicated).
The wavelength dependence of the "fast" decay stage has a smooth excitation maximum in the vicinity of 650 nm, but we were not able to correlate with it any peak in the absorption spectrum.
In any case the "fast" stage provided no more than 7-10% of the overall quantum yield.
Conclusion.
Bi-doped Mg-Al-silicate glass is developed and investigated. This glass has spectral and luminescent properties noticeably different from aluminophosphate glass but close to that of Bi, Al co-doped SiO2 glass samples prepared by MCVD method. It can be fabricated by moderate temperature routine technology. But technology upgrade is needed to produce high optical quality material. The obtained quadratic concentration dependence of Bi-caused absorption is an argument in favor of the hypothesis that the emitting centers can be associated with Bi2 dimers.
The characteristic relaxation time of 300-800 |is in combination with the high quantum yield and wide excitation spectrum makes the glass a promising material for widely tunable bulk and fiber lasers with diode or flashlamp pumping.
This investigation was supported by RFBR grant # 08-02-01054-a and by the Russian Academy of Sciences in the frames of the program "Femtosecond optics and novel optical materials".
References
1. Y.Fujimoto, M. Nakatsuka. Infrared luminescence from bismuth-doped silica glass. Jpn. J. Appl.Phys., v.40,pp. L279-L281 (2001).
2. E.Dianov, V.Dvoirin, V.Mashinskii, A.Umni-kov, M.Yashkov, A.Gurianov. CW bismuth fiber laser. Quantum Electronics, v.35, #12, pp1083-1084(2005).
3. Xian-geng Meng, Jian-rong Qiu, Ming-ying Peng et al. Near infrared broadband emission of bismuth-doped aluminophosphate glass. Optics express, v13, #5pp.1628-1634 (2005).
4. Xian-geng Meng, Jian-rong Qiu, Ming-ying Peng et al. Infrared broadband emission of bismuth-doped barium-aluminum-borate glasses. Optics express, v13, #5pp.1635-1642 (2005).
5. Ming-ying Peng, Jian-rong Qiu, Xian-geng Meng et al. Bismuth and aluminum co-doped germanium oxide glasses for super-broadband optical amplification. Optics Letters v.29, pp 1998-2000(2004).
6. Ren Jing, Chen Dan-Ping, Yang Guang, Xu Yin-Sheng, Zeng Hui-Dan, Chen Guo-Rong. Near infrared broadband emission from bismuth-dysprosium codoped chalcohalide glasses. Chin.Phys.Lett. v.24,#7, pp.1958-1960(2007).
7. B.Denker, B.Galagan, V.Osiko, S.Sverchkov, E.Dianov. Luminescent properties of Bi-doped boro-alumino-phosphate glass. Appl.Phys B 87,135-137(2007).
8. V.Sokolov, V.Plotnichenko, E.Dianov. Origin of broadband near-infrared luminescence in bismuth-doped glass. Optics Letters, v.33#13pp.1488-1490(2008).
9. S.Khonton, S.Morimoto, Y.Arai, Y.Ohishi. Luminescence characteristics of Te- and Bi-doped glasses and Glass-ceramics. J.Ceram.Soc.Jpn (v 115, 259(2007).
10. Denker B., Dianov E., Galagan B., Osiko V., Sverchkov. Spectral-luminescent properties of Bi- and Bi -Yb- doped phosphate-based glasses. ASSP-2008 technical digest, paper WE10.
11. N.Alexeev, V.Gapontsev, M.Shabotinskii, V.Kravchenko, Yu.Rudnitskii. Laser Phosphate Glasses. Nauka publishers, 1980 (in Russian). Translation: Berkeley Sci.Trans.Svc., 1983.
