Effect of oxygen impurity on the optical transmission of As 2Se3.4 glass

Article abstract of DOI:10.1023/A:1012513629483

As₂Se_3.4 glass samples with controlled oxygen content in the range (1.4-7.9) × 10^-2wt % were used to assess the effect of oxygen impurity on the IR absorption spectrum of the glass. The spectral dependence of the extinctio

Full text of DOI:10.1023/A:1012513629483

Inorganic Materials, Vol. 37, No. 11, 2001, pp. 1188–1194. Translated from Neorganicheskie Materialy, Vol. 37, No. 11, 2001, pp. 1389–1396.
Original Russian Text Copyright © 2001 by Churbanov, Shiryaev, Smetanin, Pimenov, Zaitseva, Kryukova, Plotnichenko.
Effect of Oxygen Impurity on the Optical Transmission
of As2Se3.4 Glass
M. F. Churbanov*, V. S. Shiryaev*, S. V. Smetanin*, V. G. Pimenov*,
E. A. Zaitseva*, E. B. Kryukova**, and V. G. Plotnichenko**
* Institute of High-Purity Substances, Russian Academy of Sciences,
ul. Tropinina 49, Nizhni Novgorod, 603600 Russia
** Scientific Center for Fiber Optics, Institute of General Physics, Russian Academy of Sciences,
ul. Vavilova 38, Moscow, 117756 Russia
Received June 21, 2001
Abstract—As2Se3.4 glass samples with controlled oxygen content in the range (1.4–7.9) × 10–2 wt % were used
to assess the effect of oxygen impurity on the IR absorption spectrum of the glass. The spectral dependence of
the extinction coefficient for oxygen impurity was determined in the range 600–1400 cm–1. It was shown that
the presence of 10–5 wt % O gives rise to additional losses comparable to the intrinsic losses in the CO2 lasing
range.
INTRODUCTION
and liquid. Since the distribution coefficient is
unknown, the oxygen content of the glass prepared by
melt solidification cannot be determined.
The purpose of this work was to assess the effect of
oxygen impurity on the long-wavelength optical trans-
mission of glassy As2Se3.4.
As–Se glasses are promising materials for optical
fibers [1, 2]. Glassy As2Se3 has a high mid-IR transpar-
ency. According to theoretical estimates, the minimum
optical losses in glassy As2Se3 at 6 µm are expected to
be at a level of 5 × 10–2 dB/km. Losses below 0.1 dB/m
can be achieved in the range 2–8 µm and below
0.6 dB/m at 10.6 µm.
EXPERIMENTAL
In optical measurements, we used controlled-oxy-
gen-content glasses.
Oxygen is among contaminating impurities in As–
Se glasses. When combined with glass macrocompo-
nents or gas-forming impurities, oxygen may give rise
to additional absorption bands. The effect of oxygen
impurity on the optical properties of selenium and
selenide glasses was studied in [3–8]. The extinction
coefficients of glassy selenium containing As2O3 and
SeO2 impurities were determined in [7, 8]. Demokri-
tova et al. [5] reported that the transmission spectrum of
glassy As2Se3 containing oxygen impurity showed
bands corresponding to glassy As2O3, irrespective of
whether the oxygen was introduced in the form of As2O3
or SeO2. Moynihan et al. [3] measured the absorption of
Glassy arsenic selenide was prepared by melting
high-purity arsenic and selenium at 800°C for 7 h in an
evacuated quartz tube introduced into a rocking muffle
furnace. To rule out contamination from the ambient
atmosphere, weighed amounts of arsenic and selenium
were introduced into the reactor via evaporation from
intermediate ampules in an all-welded system. The
transmission spectrum of the resultant glassy As2Se3
contained impurity absorption bands due to oxygen
combined with arsenic and selenium. In view of this,
As2Se3 was purified further by fractional vacuum distil-
lation in an open system. Glass was obtained by solidi-
fying a homogenized medium-fraction melt. According
glassy As2Se3 containing ≤1.2 × 10–2 wt % O introduced
in the form of As2O3. They found that the As2O3 was to atomic absorption data, its composition was
As2Se3.4. The absorption spectra of glass samples up to
20 mm long showed no impurity absorption in the
range 600–4000 cm–1.
To introduce oxygen into the glass, we used the fol-
lowing procedure (Fig. 1a): A massive glass sample 2
was placed in a quartz tube 1, which was connected to
the oxygen delivery line and a VMN-150 turbomolecu-
lar pump. First, the tube was pumped with valve 6
responsible for absorptions at 650, 785, 965, 1050,
1125, 1265, and 1340 cm–1. The results were inter-
preted as evidence that the glass network contained a
few oxide species. They also measured the intensity of
the 650- and 785-cm–1 bands as a function of As2O3
content, but their data cannot be used to assess the
extinction coefficient on these bands. The oxide intro-
duced into the starting charge is, in the course of melt
homogenization, redistributed between the vapor phase closed. Next, valve 5 was closed, and purified oxygen
0020-1685/01/3711-1188$25.00 © 2001 MAIK “Nauka/Interperiodica”

EFFECT OF OXYGEN IMPURITY ON THE OPTICAL TRANSMISSION OF As2Se3.4 GLASS
1189
(10–5 mol % methane and 5 × 10–4 wt % water) was fed
To the pump
from cylinder 7 into the tube by way of valve 6. Then,
the tube was sealed off at neck 3, placed in the rocking
muffle furnace, and slowly heated to 650°C. The melt
5
was homogenized at 650°C for 7 h and then cooled to
400°C, with the furnace placed in a vertical position.
Next, the tube was transferred to another furnace pre-
heated to 200°C (Fig. 1b). The tube was mounted verti-
cally so that only that part containing glass 2 was inside
the furnace 1, and the undissolved arsenic and selenium
oxides condensed at the cold end 3.
(a)
(b)
4
4
6
3
We obtained seven samples of glassyAs2Se3.4 differ-
ing in oxygen content. We also prepared control sam-
ples by the same procedure but without feeding oxygen
into the tube. In contrast to the control samples, the
glass surface in the tubes containing oxygen was
porous, as observed visually, and colorless crystals
were present at the opposite end of the tube. The x-ray
patterns of the crystals contained diffraction peaks with
d = 6.3, 3.18, 2.52, and 1.95 Å, corresponding to cubic
As2O3.
3
7
2
1
2
1
To determine the amount of oxygen that was not
incorporated into the glass, the upper part of the tube,
containing As2O3 and SeO2 deposits, was separated and
the deposits were fully dissolved in bidistilled water.
The solution was analyzed for arsenic and selenium by
atomic absorption [9]. The standard deviation in the
arsenic and selenium contents thus determined was 5%.
The amount of oxygen in the deposit was determined
under the assumption that it consisted of As2O3 and
SeO2. The amount of oxygen in the glass was found as
the difference between the amount of oxygen intro-
duced into the tube and that released into the gas phase
over the melt (Table 1). The estimated error of determi-
nation of this quantity was within 10%.
Fig. 1. (a) Experimental setup used to introduce oxygen into
the glass: (1) quartz tube, (2) massive As Se glass sam-
ple, (3) neck, (4) manometer, (5, 6) vacuum valves, (7) cyl-
inder containing high-purity oxygen. (b) Deposition of
oxides during cooling: (1) furnace, (2) glass ingot, (3) cold
end of the tube, (4) deposited oxide crystals.
2
3.4
optical finish. Transmission spectra were measured in
vacuum in the range 350–5000 cm–1 with a Brüker
For optical studies, plane-parallel plates 1, 3, and IFS-113v IR Fourier transform spectrometer and were
15 mm thick were cut and their faces were polished to then converted into absorption spectra. The spectra of
Table 1. Oxygen content of the glass samples according to atomic absorption analysis
Oxygen content
Elements in the deposit, mg
Oxygen
introducedinto glass
the tube, mg sample, g
As2Se3.4
O2 not incor-
porated into
the glass, mg
of the glass
Sample
no.
Oxides in the
deposit, mg
As
Se
wt %
at. %
1
2
3
4
5
6
7
1.55
2.74
4.16
4.86
6.33
9.20
16.00
11.12
12.08
12.17
12.31
12.02
12.15
12.00
0.03
0.09
2.8 × 10–4
5.4 × 10–4
2.9 × 10–4
2.5 × 10–4
1.7 × 10–4
1.0 × 10–4
2.2 × 10–3
0.04
0.12
0.01
0.03
0.05
0.10
2.88
3.26
6.50
1.4 × 10–2 6.7 × 10–2
2.2 × 10–2 1.1 × 10–1
3.4 × 10–2 1.6 × 10–1
3.9 × 10–2 1.9 × 10–1
2.9 × 10–2 1.4 × 10–1
4.9 × 10–2 2.4 × 10–1
7.9 × 10–2 3.9 × 10–1
0.14
0.19
0.33
0.43
9.00
11.88
13.46
26.90
10.20
20.40
INORGANIC MATERIALS Vol. 37 No. 11 2001

1190
Cm, wt %
CHURBANOV et al.
Figure 3 shows the difference absorption spectra for
glass samples and the control.
10–1
10–2
10–3
10–4
DISCUSSION
It follows from Fig. 2 that, in the range 10–2 to
10−4 wt % O, the mass spectrometry and atomic absorp-
tion data agree well.
The oxygen introduced into the tube may react not
only with arsenic and selenium but also with carbon,
hydrogen, silicon, and metallic impurities. Since the
contents of these impurities in the as-prepared glass
were at a level of 10–4 wt %, the oxygen bound to impu-
rities can be left out of consideration. As follows from
Table 1 and x-ray diffraction data, the deposit consisted
almost entirely of As2O3. Under the conditions of this
study, oxygen was likely to combine with other constit-
uents in the gas phase. Arsenic selenide dissociates
upon vaporization [11]:
10–4
10–3
10–2
10–1
Ca, wt %
As2Se3(s) = 0.0168As2Se3(g) + 0.0775As2Se3(g)
Fig. 2. Comparison of the mass spectrometry (C ) and
m
atomic absorption (C ) data on the oxygen content ofAs–Se
glasses.
a
+ 0.155AsSe(g) + 0.259As2(g)
(1)
+ 0.285As4(g) + 0.132Se2(g) + 0.396Se6(g).
samples cut from the upper and lower portions of an
ingot were identical.
Since the As–O bond (478 kJ/mol) is notably stron-
ger than the Se–O bond (419 kJ/mol) [12], oxygen is
more likely to react withAs2 andAs4 molecules. The Se
content of the deposit is two orders of magnitude lower
than the As content. The absorption spectra of the oxy-
gen-containing glasses also demonstrate that the stron-
gest impurity absorption bands are those due to arsenic
oxide species. In light of this, in interpreting the present
results we suppose that the observed changes in the
absorption spectra are mainly due to oxygen combined
with arsenic. Table 2 lists the effective Henry coeffi-
cients calculated for arsenic oxide from the data in
Table 1. The values of the Henry coefficient determined
from the equation
RESULTS
The oxygen content of the As2Se3.4 glass samples
was found to vary from 1.4 × 10–2 to 7.9 × 10–2 wt %
(Table 1). Samples 1, 2, 4, and 5 were analyzed on a
tandem laser mass reflectron [10]. Figure 2 compares
the oxygen content determined mass-spectrometrically
(Cm) and that determined by analyzing the oxide
deposit in the upper part of the tube (Ca). The contents
of carbon, hydrogen, silicon, and metals in the glass
samples were one to three orders of magnitude lower
than the oxygen content: (1–2) × 10–4 at. % C, (1–3) ×
10–3 at. % H, 10–4 wt % Si, and <10–4 wt % Al, Mg, Cu,
and Fe.
N = kp
(2)
(where N is the mole fraction of the impurity in the liq-
Table 2. Effective Henry coefficient k for As4O6
Amount of oxygen, mg
Oxygen content
As4O6 partial
pressure, kPa
Mole fraction
of As4O6 in the melt
k × 102, kPa–1
of the glass, at. %
melt
vapor
6.7 × 10–2
1.1 × 10–1
1.6 × 10–1
1.9 × 10–1
1.4 × 10–1
2.4 × 10–1
3.9 × 10–1
1.54
2.71
4.11
4.76
3.45
5.94
9.50
0.01
0.03
0.05
0.10
2.88
3.26
6.50
0.08
0.24
0.44
0.73
22
1.2
1.9
3.4
3.4
2.5
4.0
6.8
1.5
0.8
0.7
0.46
0.011
0.017
0.011
24
58
INORGANIC MATERIALS Vol. 37 No. 11 2001

EFFECT OF OXYGEN IMPURITY ON THE OPTICAL TRANSMISSION OF As2Se3.4 GLASS
1191
α, cm–1
90
80
70
60
50
40
30
20
10
0
5
7
4
6
3
2
1
900
800
700
600
500
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
6
5
4
8
8
7
2
1
4
3
3
1300
1200
1100
1000
900
Wavevector, cm–1
–2
–2
–2
–2
–2
Fig. 3. Absorption spectra of glasses containing (2) 1.4 × 10 , (3) 2.2 × 10 , (4) 2.9 × 10 , (5) 3.4 × 10 , (6) 3.9 × 10 , (7) 4.9 ×
10 , and (8) 7.9 × 10 wt % O; (1) control sample.
–2
–2
uid, and p is its partial pressure) are considered effec-
tive because, in our experiments, the distribution of
arsenic oxide between the vapor and liquid was a non-
equilibrium one. N and p were calculated for As4O6.
The spectra of the oxygen-containing glasses showed
a number of features between 400 and 1400 cm–1 that
were missing in the spectrum of the control. Their
intensities depended on the oxygen content of the glass,
while their peak positions did not.
As seen from Table 2, at a partial pressure above
0.4 kPa (in terms of As4O6), the Henry coefficient
changes markedly, and the fraction of oxygen incorpo-
rated into the glass decreases from 97–98 to 60%.
These findings suggest either a change from one oxide
species to another or, if there are a few species, changes
in their relative concentrations in the glass. Such
changes must show up in the absorption spectrum.
The contributions of individual bands to the net
absorption was evaluated by deconvoluting composite
bands as described in [7]. The shape of symmetrical
absorption bands was represented by the Voight function
I(ω) = I0(ω0)/[1 + C(ω – ω0)2 + E(ω – ω0)4],
INORGANIC MATERIALS Vol. 37 No. 11 2001

1192
CHURBANOV et al.
α, cm–1
1
101
5
8
9
3
4
100
7
2
10
11
10–1
13
16
15
14
12
10–2
1200
1000
800
600 Wavevector, cm–1
–2
Fig. 4. Deconvolution of the absorption spectrum of glassy As Se containing 1.4 × 10 wt % O into individual bands; the same
2
3.4
numbering as in Table 3.
where ω0 is the peak-intensity frequency (band posi- band width S and shape N:
tion), I0(ω0) is the maximum intensity of the band, and
S = 2{[(C2 + 4E)1/2 – C]/2E}1/2,
C and E are parameters appearing in the expressions for
N = 100/(1 + CS2 + ES4).
Table 3. Parameters of absorption bands for oxygen-con-
taining As2Se3.4 glasses
At N Ӎ 8.4, the band shape is nearly Gaussian, and
at N Ӎ 20, the band shape is nearly Lorentzian. The
shape of asymmetric bands was represented by the min-
imum possible number of symmetrical components,
whose parameters were adjusted by the differential-
moment method with subsequent minimization of the
rms deviation from the calculated absorptivities on
these bands. The error of determination of the absorp-
tion coefficient on impurity bands was 7–10%.
Frequen-
No.
N
S, cm–1
Assignment
cy, cm–1
1
2
480
482
8.9
19.8
8.4
68
2νs
8.0 As4O6, AsO3/2 (claudetite)
3
503
57
AsO3/2 (claudetite)
AsO3/2 (claudetite)
AsO3/2 (claudetite)
2νb
4
633
8.4
29
23
95
90
22
69
29
55
79
41
87
40
29
5
650
20
The parameters of the absorption bands present in
the spectra of oxygen-containing As2Se3.4 glasses are
listed in Table 3.
6
680
12.7
8.4
7
720
3νs
8
790
11
As4O6
By way of example, Fig. 4 shows the deconvoluted
9
805
10.2
8.4
AsO3/2 (claudetite)
SeO2
absorption spectrum of glassy As2Se3.4 containing 1.4 ×
10–2 wt % oxygen.
10
11
12
13
14
15
16
904
937
11.6
8.4
SeO2
The observed bands were assigned using earlier
reported spectroscopic data for arsenic oxide [13–17].
The IR-active vibrational modes of the As4O6 molecule
in the vapor phase have frequencies of 253, 346, 496,
and 782 cm–1 [13]. The absorption spectrum of arseno-
lite, consisting of As4O6 molecules, contains bands at
482, 808, 922, 938, 985, 1048, 1160, and 1265 cm–1
[16, 17]. The transmission spectrum of claudetite, a
crystalline form of arsenic oxide composed of layers of
995
Glassy As2O3
As4O6
1050
1125
1265
1340
14.5
8.4
Glassy As2O3
Glassy As2O3
Glassy As2O3
20
20
–1
Note: ν = 240 cm is the frequency of the As–Se stretch; ν is the
s
b
frequency of the As–Se–As bending.
INORGANIC MATERIALS Vol. 37 No. 11 2001

EFFECT OF OXYGEN IMPURITY ON THE OPTICAL TRANSMISSION OF As2Se3.4 GLASS
1193
α, cm–1
AsO3/2 groups, shows absorptions at 257, 285, 327,
353, 460, 500, 543, 637, 825, 865 [14], and 1100 cm–1
[16, 17]. The transmission spectrum of glassy As2O3
1
70
(a)
contains bands at 630, 830, 990, 1125, and 1340 cm–1
[14]. Selenium dioxide is characterized by absorptions
at 557, 594, 714, 904, 937, and 965 cm–1 [3, 5–7, 18].
With consideration for the changes in the transmission
spectrum upon the phase transformations of As2O3 [16,
17], the bands at 1050 and 630 cm–1 and the group of
bands at 523, 625, and 644 cm–1 can be used as the sig-
natures of arsenolite, glassy As2O3, and claudetite,
respectively.
60
50
40
30
20
10
2
3
4
0
50
Then, the bands at 482, 790, and 1050 cm–1 are due
to the cage molecule As4O6 in solution. The bands at
5
503, 633, 650, 805, 995, 1125, 1265, and 1340 cm–1 are
attributable to polymeric, amorphous, and crystalline
arsenic oxide, and the bands at 937 and 904 cm–1 are
due to the Se–O bond.
(b)
40
30
20
10
For most of the absorption bands, the intensity rises
linearly with increasing oxygen content (Fig. 5a). For
the bands at 482, 790, and 1050 cm–1, a linear rise is
observed only at low oxygen contents (Fig. 5b). The
linear portions of these curves were used to evaluate the
extinction coefficient of the corresponding absorption
band (Table 4).
6
0
0.02
0.04
0.06
0.08
The present experimental data on the effective
Henry coefficients, absorption in oxygen-containing
glasses, and intensity of absorption bands as a function
of oxygen content can be interpreted as follows: At ele-
vated temperatures in the tube, oxygen reacts with
arsenic selenide vapor to form, for the most part,
As4O6. TheAs4O6 molecules dissolved in the melt form
a layer polymer. Since the polymerization of As4O6 is
an equilibrium process, the absorption spectrum of the
glass contains bands due to both monomers and poly-
mer molecules. For this reason, the intensity of the
bands related to the As4O6 molecule no longer rises
Oxygen content of the glass, wt %
Fig. 5. Absorption coefficient as a function of oxygen con-
tent for absorption bands at (a) (1) 633, (2) 650, (3) 805,
–1
(4) 503, (b) (5) 790, and (6) 482 cm
.
The largest extinction coefficient in the range of the
linear variation of intensity with oxygen content is
1270 cm–1/wt % (on the 790-cm–1 band). In As2Se3.4
Table 4. Linear (εl) and integral (εs) extinction coefficients
of oxygen-related absorption bands in the spectra of glassy
upon a further increase in oxygen content. The disor- As2Se3.4
dered continuousAs2O3 polymer network formed in the
Maximum
εl, cm–1/wt %
εs, cm–2/wt %
melt may undergo ordering upon solidification, produc-
ing regions with the claudetite structure. A fraction of
the bonds broken upon As4O6 cage opening may inter-
act with As–Se groups of the melt to form Se–O bonds.
Since the As–O bond is notably stronger than the Se–O
bond, As4O6 homopolymerization is energetically more
favorable than copolymerization with arsenic selenide.
For this reason, the proportion of Se–O bonds in the
glass is lower than that of As–O, in agreement with the
relative intensities of the corresponding absorption
bands (Fig. 3).
frequency, cm–1
633
650
860
380
280
190
7
28130
13500
22200
12160
660
805
503
995
1125
1265
1340
790
11
1080
460
7
1
50
The extinction coefficients of the observed absorp-
tion bands were used to construct the spectral depen-
dence of the extinction coefficient for oxygen impurity
in glassy As2Se3.4 over the entire range of oxygen con-
tents studied (Fig. 6).
1270
540
30
32370
6390
1660
710
482
1050
937
11
INORGANIC MATERIALS Vol. 37 No. 11 2001

1194
ε(λ), cm–1/wt %
CHURBANOV et al.
REFERENCES
1. Dianov, E.M., Petrov, M.Yu., Plotnichenko, V.G., and
Sysoev, V.K., Evaluation of Minimal Transmission
Losses in Chalcogenide Glasses, Kvant. Elektron. (Mos-
cow), 1982, vol. 9, no. 4, pp. 798–800.
2. Plotnichenko, V.G., Current and Potential Application
Areas of IR Optical Fibers, Vysokochist. Veshchestva,
1994, no. 4, pp. 34–42.
3. Moynihan, C.T., Macedo, P.B., Maklad, M.S., et al.,
Intrinsic and Impurity Infrared Absorption in As2Se3
Glass, J. Non-Cryst. Solids, 1975, vol. 17, pp. 369–385.
4. Ma, D.S., Danielson, P.S., and Moynihan, C.T., Bulk and
Impurity Infrared Absorption in 0.5As2Se3–0.5GeSe2,
J. Non-Cryst. Solids, 1980, vol. 37, no. 4, pp. 181–190.
103
102
101
100
1400
1200
1000
800
600
Wavevector, cm–1
5. Demokritova, N.V., Vinogradova, G.Z., Bel’skii, N.K.,
and Lopatto, Yu.S., Removal of Oxygen from Selenium,
Izv. Akad. Nauk SSSR, Neorg. Mater., 1984, vol. 20,
Fig. 6. Spectral dependence of the extinction coefficient for
oxygen impurity in glassyAs Se ; (1.4–5.0) × 10 wt % O.
–2
2
3.4
no. 3, pp. 511–514.
^
6. Le al, D. and Srb, I., Synthesis of Chalcogenide Glasses
z
of the Se–Ge and Se–As Systems without Traces of Oxy-
gen, Collect. Czech. Chem. Commun., 1973, vol. 36,
pp. 2091–2097.
glass with optical losses at 920–1090 cm–1 (CO2 lasing
range) at a level of intrinsic losses, the oxygen content
should be no higher than 10–5 wt % O. The present data
on the extinction coefficient of oxygen impurity can be
used to assess the oxygen content of As–Se glasses
from IR spectrometric data.
7. Bychkova, T.I., Vinogradova, G.Z., Voitsekhovskii, V.V.,
and Plotnichenko, V.G., ImpurityAbsorption ofAs2O3 in
Glassy Selenium, Vysokochist. Veshchestva, 1990, no. 4,
pp. 203–207.
8. Voight, B. and Dresler, G., Bestimmung undAbtrennung
von Sauerstoffverunreinigungen in Reinst-Selen, Anal.
Chim. Acta, 1981, vol. 127, pp. 87–92.
9. Malyshev, A.Yu., Pimenov, V.G., and Zaitseva, E.A.,
Determination of Arsenic, Antimony, and Selenium
Impurities in High-Purity Sulfur, Anal. Kontrol, 2000,
vol. 4, no. 4, pp. 329–333.
10. Kovalev, I.D., Malyshev, K.N., and Shmonin, P.A., Tan-
dem Laser Mass Reflectron for Determination of Gas-
Forming Impurities in Solids: Design and Concept, Zh.
Anal. Khim., 1998, vol. 53, no. 1, pp. 38–42.
11. Gorbov, S.I. and Krestovnikov, A.N., Vapor Pressure of
Arsenic Selenide, Zh. Neorg. Khim., 1968, vol. 13, no. 6,
pp. 1482–1487.
12. Vedeneev, V.I., Gurvich, L.V., Kondrat’ev, V.N., et al.,
Energiya razryva khimicheskikh svyazei: Spravochnik
(Energy of Bond Breaking: A Handbook), Moscow:
Akad. Nauk SSSR, 1961.
13. Behrens, R.G. and Rosenblatt, G.M., Vapor Pressure and
Thermodynamics of Octahedral Arsenic Trioxide (Arse-
nolite), J. Chem. Thermodyn., 1972, no. 4, pp. 175–190.
14. Papatheodorou, G.N. and Solin, S.A., Vibrational Exci-
tations of As2O3: I. Disordered Phases, Phys. Rev. B:
Solid State, 1976, vol. 13, no. 4, pp. 1741–1751.
CONCLUSION
Our results demonstrate that oxygen impurity is
responsible for a large number of absorption bands in the
spectrum of glassy As2Se3.4. It is shown that most of the
observed bands can be identified with those in the absorp-
tion spectra of As4O6 molecules, glassy As2O3, and clau-
detite. Two bands arise from Se–O vibrations. The data on
the observed absorption bands, intensity of absorption
bands as a function of oxygen content, and effective Henry
coefficients for As4O6 suggest that As4O6 cage opening
leads to the formation of an amorphous polymer capable
of crystallizing upon melt solidification. The spectral
dependence of the extinction coefficient of oxygen impu-
rity in the range 7–20 µm was established. Oxygen impu-
rity was shown to have a significant effect on the transpar-
ency of glassyAs2Se3.4. The presence of 10–5 wt % O gives
rise to additional losses comparable to the intrinsic losses
in the spectral range 9–11 µm.
15. Edmond, J.T. and Redfearn, M.W., Infra-Red Study of
As2Se3-Type Glasses in the Wavelength Range 1.25 to
25 µm, Proc. Phys. Soc. (London), 1963, vol. 81,
ACKNOWLEDGMENTS
pp. 378–382.
We are grateful to I.D. Kovalev and D.A. Ovchinni-
kov for performing the mass spectrometric analyses.
We also acknowledge the assistance of N.D. Malygin
with x-ray diffraction analyses.
^
16. Lezal, D. and Srb, I., Studium der Ultrarotspektren ver-
schiedener Modifikationen der Arsen(III)-Oxids, Chem.
Zv., 1971, vol. 25, pp. 32–43.
^
^
17. Lezal, D. and Konák, K., The Characterization of the
Infrared Absorption Spectra of the Vitreous, Cubic, and
Monoclinic Modifications of As2O3, J. Non-Cryst. Sol-
ids, 1995, vol. 192/193, pp. 187–190.
This work was sponsored through the Support to Lead-
ing Scientific Schools Program, grants no. 00-15-99008
and no. 00-15-96650.
INORGANIC MATERIALS Vol. 37 No. 11 2001