Physical, structural and optical characterization of silicate modified bismuth-borate-tellurite glasses

Physical, structural and optical characterization of silicate modified bismuth-borate-tellurite glasses

Accepted Manuscript Physical, structural and optical characterization of silicate modified bismuth-boratetellurite glasses Neelam Berwal, Sunil Dhankh...

3MB Sizes 1 Downloads 94 Views

Accepted Manuscript Physical, structural and optical characterization of silicate modified bismuth-boratetellurite glasses Neelam Berwal, Sunil Dhankhar, Preeti Sharma, R.S. Kundu, R. Punia, N. Kishore PII:

S0022-2860(16)30863-8

DOI:

10.1016/j.molstruc.2016.08.033

Reference:

MOLSTR 22859

To appear in:

Journal of Molecular Structure

Received Date: 8 March 2016 Revised Date:

11 August 2016

Accepted Date: 11 August 2016

Please cite this article as: N. Berwal, S. Dhankhar, P. Sharma, R.S. Kundu, R. Punia, N. Kishore, Physical, structural and optical characterization of silicate modified bismuth-borate-tellurite glasses, Journal of Molecular Structure (2016), doi: 10.1016/j.molstruc.2016.08.033. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Physical, Structural and Optical Characterization of Silicate Modified Bismuth-Borate-Tellurite Glasses

a

RI PT

Neelam Berwal a, Sunil Dhankhara, Preeti Sharmaa, R. S. Kundu a, R. Puniaa* and N. Kishore b Department of Applied Physics, Guru Jambheshwar University of Science & Technology, Hisar-

b

SC

125001

Department of Physics, Central University of Haryana, Mahendergarh-123029

M AN U

*Corresponding author’s e-mail: [email protected]

Abstract

The quaternary glass system xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 has been prepared by

TE D

melt-quench technique. The amorphous nature of glass samples has been ascertained by X-ray diffraction patterns. The variations in density, molar volume and crystalline volume with glass compositions have been discussed. A non-linear change has been observed in glass transition

EP

temperature and optical band gap energy. Raman and FTIR spectral studies suggest that glass network is mainly built up of BO3, BO4, SiO4, and TeO3 structural units, whereas BiO3 exists as

AC C

both network modifying [BiO6] octahedral as well as network forming [BiO3] pyramidal structural units. The values of optical band gap energy have been estimated from fitting of both Mott and Davis’s model and Hydrogenic excitonic model (HEM) with experimental data of absorption spectra. The HEM model shows good agreement with experimentally observed absorption spectra, which indicates the exciton formation in studied glass system. The non-linear compositional change in optical band gap energy is related with the structural changes occurring

1

ACCEPTED MANUSCRIPT

in present glass samples. The Urbach energy has also been estimated. The range of metallization criterion suggests that prepared glasses may be considered as new nonlinear optical materials.

RI PT

Keywords: Bismuth silicate glass; DSC; Raman; FTIR; UV-Vis spectroscopy; HEM.

1. Introduction:

Recently, structural and optical properties of Bi2O3 based oxide glasses have been studied

SC

vastly [1-9] because of its role as both network former and modifier. These glasses have interesting application as transmitting windows in IR region, photocatalytic degradation

M AN U

materials, thermal and mechanical sensors and layers for optical and electronic devices [5, 9-12]. B2O3 is one of the basic glass formers and mainly exists in the form of trigonal BO3 and tetrahedral BO4 units. It forms glass at lower temperature with good transparency, high chemical durability and thermal stability [13, 14]. Bismuth borosilicate glasses are important in the present

TE D

because of their useful luminescent and non-linear optical properties [12, 15-16]. Tellurite glasses are important due to their promising optical properties such as high linear and non-linear refractive indices, low melting point and good transmission in IR regions [5, 17- 18].

EP

Silicate glasses are chemically durable, thermally stable and optically transparent at excitation and lasing wavelength [19]. In addition, high viscosity of these glasses allows the

AC C

glass to be formed, cooled and annealed without crystallization. They are useful in optics as lenses or beam splitters, in optical telecommunications, micro- and optoelectronics and in near IR-windows [20-21]. A review of literature shows that there are many reports on B2O3: Bi2O3: SiO2 [22], TeO2: Bi2O3: SiO2 [18], Bi2O3: SiO2: Fe2O3 [4], TeO2: Bi2O3: B2O3: ZnO [5] and TeO2: Bi2O3: B2O3: PbO [17] glass systems. But there is hardly one report on Bi2O3: B2O3: TeO2: SiO2 [8] glass system, where we studied effect of B2O3 on Bi2O3: TeO2: SiO2 glass system

2

ACCEPTED MANUSCRIPT

and noticeable change in physical, structural and optical properties has been observed on addition of borate to bismuth tellurite silicate glass system. To optimize Bi2O3: B2O3: TeO2: SiO2 glass system for use as non-linear optical material

RI PT

and in order to explore its technological applications, the effect of SiO2 on physical, structural and optical properties of bismuth-borate-tellurite glasses have been studied and reported in

SC

present manuscript.

2.1 Glass Preparation:

M AN U

2. Experimental Details:

Glass samples of compositions xSiO2-(80-x) Bi2O3-15B2O3-5TeO2, with x=0, 5, 10, 15, 20, 25 and 30 were prepared by conventional rapid melt-quenching technique using analar grade

TE D

chemicals H3BO3, TeO2, Bi2O3 and SiO2 . The raw materials in appropriate molecular ratios were thoroughly mixed in an agate pestle mortar to form batches of 20 g. The batch mixture was then melted in a porcelain crucible in an electrically heated muffle furnace under ordinary

EP

atmospheric conditions at a temperature of about 1100˚C which was kept for 1 hr and the melt

AC C

was occasionally swirled to ensure proper mixing, homogeneity and to avoid the presence of bubbles. Finally, the glasses were formed by pouring the melt at room temperature on to a stainless steel block and it was quickly quenched by pressing with another stainless steel block. The as prepared samples were annealed at 350°C (below its glass transition temperature) for four hours to minimize the internal mechanical stress (which may have developed during the quenching process) and subsequently cooled to room temperature in furnace.

2.2 Glass Characterization: 3

ACCEPTED MANUSCRIPT

The amorphous nature of glass samples was ascertained by X-ray diffraction patterns using Rigaku Table-Top X-ray Diffractometer with Cu-K radiation (λ=1.54Å) at 30 KV and 15 mA in 2θ range 20°-80° with a scanning rate of 4° per minute. The density (ρ) measurements

RI PT

were made by the Archimedes’ method using SI-234 Denver Instrument with xylene as buoyant liquid. The glass transition temperature (Tg) of samples was measured by differential scanning calorimetry (DSC) technique using TA (thermal analysis and analyzers) instrument, model

SC

No.Q600 SDT, with a heating rate of 20˚C/min in temperature range 200°C - 800°C in nitrogen atmosphere. The IR spectra of powdered glass samples were recorded at room temperature using

M AN U

KBr pellet technique on a Shimadzu IRaffinity-I 8000 FTIR spectrophotometer in wavenumber range 400 cm-1 -1500 cm-1 with spectral resolution of 2 cm-1. The powdered glass samples were thoroughly mixed with dry KBr in a ratio 1:20 by weight and then pellets were formed under a pressure of 109 Pa to produce clear homogenous discs. The room temperature Raman

TE D

measurements of the polished samples were performed on Renishaw Invia Reflex Micro Raman Spectrometer with Ar ion laser (514 nm) under back scattering configuration in the range 100 cm-1 -1700 cm-1 with spectral resolution of 1 cm-1. Laser power at the sample was ~ 100 mW and

EP

typical acquisition time was 30 sec. FTIR and Raman spectra were found to be consisting of a group of broad bands that are composed of a number of overlapping peaks. To resolve these

AC C

peaks and to find out the expected exact position of peaks, deconvolution of FTIR and Raman spectra was performed using “multiple peaks fit” module of Origin Pro 8.6 software by using Gaussian function. The quality of Gaussian fitting is evaluated by best fit parameter R2 and in present case R2 is greater than 0.9999. The statistical validation has also been tested by applying F-test following the procedure reported earlier [3, 6, 23]. The optical absorption spectra of prepared glass samples were recorded at room temperature using 3100 MC Shimadzu UV-Vis.

4

ACCEPTED MANUSCRIPT

spectrophotometer in the wavelength range 200 nm -800 nm with spectral resolution of 1 nm and spectral slit width of 2 nm.

RI PT

3. Results and Discussion: 3.1. XRD Analysis:

SC

X-ray diffraction (XRD) patterns of prepared glass samples presented in Fig.1 exhibit broad diffuse scattering around low angle region, instead of sharp crystalline peaks, which

M AN U

ensure the amorphous nature as well as absence of long range atomic arrangement in studied glass samples.

32. Density, Molar Volume and equivalent Crystalline Volume:

TE D

The density for all glass samples was measured using Archimedes’ principle and corresponding molar volume (Vm) of samples was calculated using the relation [1]: Vm = ∑ xi M i / ρ

(1)

EP

where xi , Mi are respectively the molar fraction and molecular weight of the ith component and ρ is the density of the sample. Their values are listed in Table-1. Perusal of data presented in

AC C

Table-1 show that density and molar volume of the samples decrease from 5.995 g/cm3 to 5.697 g/cm3 and 65.06 cm3/mole to 47.09 cm3/mole, respectively with increase in SiO2 content at expense of Bi2O3 content. The decrease in density is an expected result as heavier Bi2O3 (atomic mass 465.95) is being replaced by lighter SiO2 (atomic mass 60.09). The decrease in molar volume with increase in SiO2 content is mainly attributed to smaller radii and bond lengths of SiO2 than that of Bi2O3 and thus results in a shrinking of free volume [24]. Further, according to

5

ACCEPTED MANUSCRIPT

Pan et al. [25], SiO2 tetrahedral can be incorporated in the network of more flexible structure of bismuth-oxygen polyhedral, [BiOn] due to more ionic nature of Bi-O bond and hence the decrease in molar volume in presently studied glasses.

volume (Vc) of the glass samples is given by the formula [1]: Vc = ∑ xiVi

RI PT

The corresponding molar volume for their crystalline phases or equivalent crystalline

(2)

SC

where Vi is molar volume of the ith component in crystalline phase. The calculated values of Vc

M AN U

are included in Table-1. The data of molar volume and equivalent crystalline volume are compared in Fig.2, which in turn shows that values of Vm of glasses for different compositions are greater than the corresponding values of the Vc, indicating that structure of non-crystalline counter part of these mixed crystalline phases are expanded with some excess structural volume.

TE D

3.3 Differential Scanning Calorimetry (DSC):

The glass transition temperature (Tg) has been estimated from minima occurred in the graph of dH/dT versus temperature for all glass samples, as shown in Fig.3. Oxygen packing

EP

density (OPD), which is a measure of the tightness of packing of oxide network, was calculating

AC C

using following relation [2, 26]:

OPD = 1000(ρ / M )n

(3)

where M is molecular weight, ρ is density and n is number of oxygen atoms per formula unit. The observed values of OPD and Tg are presented in Table-1. Fig.4 shows variation of Tg and OPD with concentration of SiO2. Perusal of data presented in figure shows that glass transition temperature exhibits a nonlinear behavior with increase in SiO2 content. This type of behavior in Tg may be due to structural change occurring in studied glass samples. 6

ACCEPTED MANUSCRIPT

As glass transition temperature is strictly related to density of cross-linking and tightness of packing in the network. Therefore, increase in Tg with increase in SiO2 content (for x ˂ 20) is due to the replacement of weak Bi-O-Bi cross-linkages by stronger Si-O-Si cross linkages which

RI PT

in turn increases the cohesiveness of network and hence the glass network structure. However, further decrease in Tg for x ≥ 20 is due to the dominating role played by the network modifying

SC

BiO6 octahedral units (also supported by structural analysis) [1].

M AN U

3.4 Fourier Transform Infrared (FTIR) Spectroscopy:

The effect of substitution of unconventional glass former (Bi2O3) by conventional glass former (SiO2) on structural properties of xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass compositions has been investigated by recording their FTIR spectra. FTIR spectra of the prepared glass samples recorded at room temperature in the spectral range 400 cm-1 - 1500 is shown in Fig. 5(a).

TE D

The recorded FTIR spectra is characterized by superimposed broad peaks, which are deconvoluted into distinct peaks using Gaussian distribution in order to find out exact peak

EP

position corresponding to structural units present in glass matrix. The typical deconvoluted FTIR spectra (for the sample with x=25) is shown in Fig. 5(b) The peak position (xc), amplitude (A)

AC C

and full width at half maximum (w) for all the peaks observed from deconvoluted FTIR spectra are presented in Table-2.

The well distinguished peak observed between 449 cm-1 and 476 cm-1 in IR spectra is due

to BiO6 octahedral units [1]. The absorption peak in the region 505 cm-1 - 556 cm-1 may be attributed to Bi-O and Bi-O-Bi stretching vibrations of [BiO6] octahedral structural units [27, 28]. Another IR peak between 697 cm-1 -711 cm-1 observed in all glass samples may be due to symmetric stretching vibrations of Bi-O bond in BiO3 pyramidal units [1] and/or presence of Te7

ACCEPTED MANUSCRIPT

O vibrations in trigonal pyramid [TeO3] groups [29]. The intensity of this peak remains almost same as per the expectations, as TeO2 content remains same in all the compositions. The peak observed at around 832 cm-1 in IR spectra is assigned to symmetric stretching

RI PT

vibrations of Bi-O bond in BiO3 group [1, 3]. The IR peak at ~ 870 cm-1 in all studied glass samples is attributed to BiO6 octahedral structural units [8, 23, 30]. Intensity of this peak is found to be decreases with increase in x up to x=15 but increases in glass samples with x ≥ 20,

SC

indicating that number of BiO6 octahedral structural units increases for these compositions (x ≥ 20). IR peak in the region 968 cm-1 - 1025 cm-1 is observed in all glass samples may be assigned

M AN U

to stretching vibrations of B-O-Bi linkages. The IR peak observed at around 1090 cm-1 for glass sample with x=0 may be attributed to B-O stretching vibrations of BO4 units in tri-, tetra- and penta-borate structural units [23, 31]. IR peak observed between 1108 cm-1 -1136 cm-1 in all compositions except for the sample with x=0 is due to asymmetric stretching vibration of Si-O

TE D

bonds in SiO4 tetrahedral units and/or linkage vibrations of Bi-O-Si and Bi-O-Bi [3]. The position and intensity of this band increases on wave number scale with increase in SiO2 content. IR peak between 1205 cm-1 - 1220 cm-1 observed in all compositions are due to stretching

EP

vibrations of B-O bonds of trigonal [BO3] groups [32]. IR peak in the spectral range 1243 cm-1 - 1360 cm-1 is observed in all glass samples

AC C

indicate the presence of asymmetric stretching vibration of BO3 triangular units in meta-borate, pyro-borate and ortho-borate groups [32]. Another IR peak between 1505 cm-1 - 1538 cm-1 can be ascribed to stretching vibrations of BO3 triangular units attached to large segments of borate network [23].

3.5 Raman Spectroscopy: 8

ACCEPTED MANUSCRIPT

The Raman spectra of xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glasses system taken at room temperature in spectral wave number range 100 cm-1 -850 cm-1 and 850 cm-1 - 1700 cm-1 are shown in Figs. 6(a) and (b) respectively. The recorded Raman spectra are characterized by

RI PT

several broad bands which may be the mixture of more than one peak. For better identification of these peaks, Raman spectra is deconvoluted and typical deconvoluted spectra in the spectral range 100-850 cm-1 and 850-1700 cm-1 (for the sample with x=20) are shown in Figs. 6(c) and

SC

(d), respectively. Each of the deconvoluted spectra exhibits several peaks and peak position (xc), amplitude (A) and full width at half maximum (W) of these peaks are presented in Table-3.

M AN U

The most dominating peaks in Raman spectra observed at around 113 cm-1 and 150 cm-1 in all glass compositions, are related to heavy metal ion vibration i.e. vibration involving motion of Bi+3 cations in BiO6 octahedral and/or BiO3 pyramidal units [4, 7, 8, 24]. The intensity of first peak decreases slightly as SiO2 content increases, indicating that number of Bi+3 cations in BiO6

TE D

and BiO3 units decreases with increase in SiO2 content. The broad band in present Raman spectra observed ~ 295 cm-1 can be attributed to Bi-O-Bi stretching vibrations in distorted BiO6 octahedral units [27, 33].

EP

A well distinguished peak centred at around 422 cm-1 in Raman spectra can be assigned to the Bi-O-Bi vibrations in both BiO3 pyramidal and BiO6 octahedral structural units [5]. This

AC C

peak is almost of constant intensity and full width at half maximum. The Raman peak between 520 cm-1 - 570 cm-1 belongs to the Bi-O¯ stretching vibrations (i.e. vibrations of bismuth associated with non-bridging oxygens (NBOs)) of BiO6 octahedral units [8, 24, 34]. The Raman intensity of this peak remains almost constant with increase in x up to x=15, thereafter it increases slightly with increase in SiO2 content, indicating that BiO6 group containing NBO increases for glass samples with x ≥ 20. The Raman band centred at around 740 cm-1 is related to

9

ACCEPTED MANUSCRIPT

the stretching mode of TeO3 trigonal units [5, 8]. Iintensity of this peak remains almost same as per expectations, because TeO2 content remains same in all the compositions.

RI PT

The appearance of another Raman peak at 950 cm-1 for glass sample with x=0 may be attributed to B-O stretching vibrations of BO4 units in tri-, tetra- and penta-borate structural units [32]. Raman peak between 1000 cm-1 - 1109 cm-1 observed in all compositions except with x=0

SC

is assigned to symmetric stretching vibration of SiO4 tetrahedral units with 2 NBOs [4]. With increase in silicate content, this peak shifted towards high wavenumber side. Raman peak

M AN U

between 1156 cm-1 -1208 cm-1 observed in all compositions are due to the stretching vibrations of B-O bonds of trigonal [BO3] groups [23].

Another Raman peak between 1315 cm-1 - 1410 cm-1 is observed in all glass samples indicate the presence of asymmetric stretching vibration of BO3 triangular units in meta-borate,

1

TE D

pyro-borate and ortho-borate groups [23]. In Raman spectra, peak between 1481 cm-1 - 1640 cmcan be ascribed to stretching vibrations of BO3 triangular units attached to large segments of

borate network [32].

EP

Hence, taking in to consideration these structural changes of FTIR and Raman spectra, it is observed that increase of SiO2 content in glass structure leads to following: network former

AC C

SiO2 is present in SiO4 tetrahedral structural units with two NBOs per silicon. Bi2O3 plays role of both network modifier BiO6 octahedral as well as network former BiO3 pyramidal units for all glass samples but for glass samples with x ≥ 20, number of BiO6 octahedral structural units increases with increase in x (SiO2 content). B2O3 exists in the form of BO3 trigonal and BO4 tetrahedral structural units with B-O-Bi linkages and TeO2 in the form of TeO3 structural units.

10

ACCEPTED MANUSCRIPT

3.6 Optical Spectroscopy: Optical absorption spectra of all glasses have been recorded at room temperature in UV-

α( ν) = A / t

RI PT

Vis region. The absorption coefficient was determined using following relation [35]: (4)

where A is absorbance and t is thickness of prepared samples. The absorption coefficient α (ν) (as

SC

a function of photon energy) in many amorphous materials reflects the density of states at the band edges as proposed by Mott and Davis and is given by [36, 37]:

where

(5)

M AN U

αhν = B (hν − Eg 1 ) r

is absorption coefficient, hν is incident photon energy. B is energy independent

constant , Eg1 is optical band gap energy and index r is a constant that determines type of optical transitions and can have different values, viz. 1/2, 2, 1/3 and 3 corresponding to direct allowed,

TE D

indirect allowed, direct forbidden and indirect forbidden transitions, respectively [1]. The values of optical band gap energy (Eg1) have been estimated from the fitting of experimentally observed absorption coefficient data with Eq. (5). ). A typical graph for glass sample with x=15 showing

EP

fitting of experimentally observed optical absorption coefficient data with Mott-Davis model for r = 1/2, 2, 1/3 and 3 has been presented in Fig. 7. Perusal of data presented in Fig. 7, both r = 2

AC C

and 3 fit good with experimental data, suggesting indirect band transitions in studied glass system. Fitting of Eq. (5) with experimental data of all studied glass compositions has been presented in Fig. 8 (a) and (b) for r = 2 and 3, respectively. The values of optical band gap (Eg1) for indirect allowed and forbidden transitions obtained from fitting are presented in Table 1. Perusal of the data presented in Table 1, shows that Eg1 increases with increase in SiO2 content up to x = 15 and thereafter it decreases. This nonlinear variation in Eg1 may be associated with the structural changes occurring in studied glass samples. The increase in Eg1 with increase in 11

ACCEPTED MANUSCRIPT

SiO2 content corresponding to x = 15 is due to the contraction and strengthening of glass network, illustrated by decrease in molar volume or by the replacement of lower bond strength of Bi-O bond (80.3 Kcal/mole) by higher bond strength of Si-O bond (110 Kcal/mole) [38].

RI PT

However, further decrease in Eg1 may be due to the dominating role played by network modifying BiO6 structural units over network forming BiO3 structural units. This is evident from

SC

structural studies using FTIR spectra.

Mott and Davis’s model neglects the contributions of electron-hole pair (exciton) or

M AN U

electron-impurity interactions. As the formation of bounded electron-hole pair increases the optical transition rate. Hence, to study contribution of excitionic transitions for studied glass samples, Hydrogenic Excitonic Model (HEM) has been applied on absorption spectra. Shape of absorption spectra of studied glass samples may be explained using (HEM), which provides an explicit expression for optical absorption due to bound and continuum excitons [5, 39]. Excitons

TE D

play an important role in optical properties of semiconductors. Intrinsic parameters such as band gap and excitonic binding energy can be determined by fitting the absorption spectrum with an appropriate analytical model. Absorption coefficient (α) introduced by Elliott - Washington

  hω − Eg 2  ∞ R C0R1/2  ∞ 2R 1 π sinh(2u+ ) Γm Γc π + + arctan − + ∑ 3    ∑ 2   2 3 + − 2 E m=1 m ( E − Em ) + Γ m2 2  2  Γ c  m=1 m ( E − Em ) + Γ c 2 cosh(2u ) − cos(2u ) 

AC C

α ( E) =

EP

model for hydrogenic excitons is given as:

(6)

with 1/2

2 2 1/2 1/2    R   ( E − E g 2 ) + Γ c  ± ( E − E g 2 )  ± u =π   ( E − E g 2 )2 + Γ c 2  2   

12

ACCEPTED MANUSCRIPT

Em = E g 2 −

Γ m = Γ c − ( Γ c − Γ1 ) / m 2 ,

m = 1, 2, 3.

RI PT

and

R m2

where Co is absorption strength parameter, Eg2 is optical band gap energy, E is incident photon energy, R is excitonic binding energy, and Г1 and Гc are linewidths of the m = 1 state and of

SC

continuum, respectively. The first term near the band edge gives rise to a single peak, which is centred at energy Em = 1 and approximately of width Г1. Rest of the terms represent excitonic

M AN U

continuum and give a step like absorption edge of width Гc above the band gap. The fitting of experimental data fitted Hydrogenic excitonic model (Eq.6) is shown in Fig. 9 and values of parameters Eg2, R, Г1, Гc and Co obtained from fitting are listed in Table-4. Values of χ2 and R2 corresponding to fitting of Eq. (6) to experimental data are also included in Table-4. Best fit

TE D

parameter R2 has values more than 0.99 indicating thereby good agreement between experimental data and theoretical model. So, optical transitions in prepared glass system are of hydrogenic excitonic type. The values of fitting parameters R, Г1, Гc and Co are in good

EP

agreement with values reported elsewhere [3, 6, 8, 23]. Hydrogen excitionic model considers direct optical transitions. As in amorphous solids, both direct and indirect transitions are

AC C

possible due to lack of long range order. So, the hydrogenic excitionic model helps to obtain the direct optical band gap values [23]. The relation between Urbach energy ∆E and absorption coefficient α (ν) is given by well

known Urbach law [1, 40]:

α( ν) = B exp(hν / ∆E )

13

(7)

ACCEPTED MANUSCRIPT

where B is constant and ∆E is Urbach energy, which is usually interpretted as the width of band tails of localized state in the band gap. The relation can be rewritten as:

ln α(ν) = hν / ∆E + constant

RI PT

(8)

The values of Urbach energy (∆E) were calculated by determining slope of linear regions of lnα verses hν curve (called Urbach plots and are shown in Fig. 10) and taking their

SC

reciprocals. The values of ∆E thus determined are presented in Table 1. The materials which have large value of ∆E would have great tendency to convert weak bond into defects and

M AN U

reducing the long range order. For present glass samples, values of ∆E are in the range 0.17-0.22 eV which corresponds to amorphous semiconductors as reported by E. Davis and N. Mott [36] and it is found to decrease with increase in SiO2 content which in turn decreases the fragility nature of the glass network.

TE D

The molar refractivity (Rm) is calculated by using the relation [41, 42]:

Vm [1 − E g / 20] = Rm

(9)

EP

The molar polarizability (αm) is directly proportional to molar refractivity of the material

AC C

and is given by the relations [43]:



3 

αm =   Rm  4Π N 

(10)

where N is Avogadro’s number. The values of αm and Rm are listed in Table-1. The metallization criterion (M) of an oxide, which gives us information about non-

metallic nature of solids on the basis of its band gap energy, is given [5] as:

M = 1 − Rm / Vm

14

(11)

ACCEPTED MANUSCRIPT

Oxide glasses with good optical non-linearity are reported to have metallization criterion in the range of 0.30-0.45 [44]. The calculated value of metallization criterion for present glass

considered as new non-linear optical materials [5, 8].

4. Conclusions:

RI PT

samples (listed in Table-1) is in the range 0.33-0.36. Hence, material under investigation may be

SC

Bismuth silicate glasses with compositions xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 have been successfully prepared by melt-quenching technique. Presence of broad band in XRD patterns

M AN U

demonstrates amorphous nature of glass samples. The density, molar volume and crystalline volume decrease with increase in SiO2 content. From analysis of Raman and FTIR spectra of these glasses, it was found that glass network is built up of mainly BO3, BO4, SiO4, and TeO3 structural units. Bi+3 cations are incorporated in glass network as network modifying [BiO6]

TE D

octahedral as well as network forming [BiO3] pyramidal units for all glass compositions, however, for glass samples with x ≥ 20, number of BiO6 octahedral structural units increases with increase in SiO2 content. The deviation from linearity in various parameters such as glass

EP

transition temperatures and optical band gap is due to the dual role (glass former and glass modifier) played by Bi+3 ions in glass structure. Optical band gap energy calculated from fitting

AC C

of Mott and Davis’s model with experimentally observed absorption spectra show nonlinear behaviour, whereas molar refractivity and molar polarizability increase with increase in SiO2 content. The values of direct optical band gap, excitonic binding energy and linewidths of m = 1 state and of continuum have also been obtained from fitting of experimental data with Hydrogenic excitonic model. The HEM model shows good agreement with the experimental

15

ACCEPTED MANUSCRIPT

data, which indicates the excitonic formation in studied glass system. The presently studied glass samples may be considered as potential candidates for nonlinear optical applications.

RI PT

Acknowledgements

Authors are thankful to University Grants Commission (UGC), New Delhi and

SC

Department of Science and Technology (DST-FIST) and DRDO-IRDE Dehradun for providing

AC C

EP

TE D

M AN U

financial support.

16

ACCEPTED MANUSCRIPT

References: [1] R. Punia, R. S. Kundu, J. Hooda, S. Dhankhar, S. Dahiya, N. Kishore, J. Appl. Phys. 110

RI PT

(2011) 033527(1-6). [2] S. Dahiya, R. Punia, S. Murugavel, A. S. Maan, J. Mol. Struct. 1079 (2015) 189-193.

SC

[3] R. S. Kundu, M. Dult, R. Punia, R. Parmar, N. Kishore, J. Mol. Struct. 1063 (2014) 77-82.

M AN U

[4] R. Parmar, R. S. Kundu, R. Punia, P. Aghamkar, N. Kishore, Physica B 450 (2014) 39-44. [5] R. S. Kundu, S. Dhankhar, R. punia, K. Nanda, N. Kishore, J. Alloys Comp. 587 (2014) 6673.

[6] L. Singh, V. Thakur, R. Punia, R. S. Kundu, A. Singh, Solid State Sci. 37 (2014) 64-71.

TE D

[7] M. Dult, R. S. Kundu, N. Berwal, R. Punia, N. Kishore, J. Mol. Struct. 1089 (2015) 32-37. [8] N. Berwal, R. S. Kundu, K. Nanda, R. Punia, N. Kishore, J. Mol. Struct. 1097 (2015) 37-44.

AC C

(2015) 197-203.

EP

[9] V. Thakur, H. S. Kushwaha, A. Singh, R. Vaish, R. Punia, L. Singh, J. Non-Cryst. Solids 428

[10] D. W. Hall, M. A. Newhouse, N. F. Borelli, W. H. Dumbaugh, D. L. Weidman, Appl. Phys.

Lett. 54 (1989) 1293.

[11] C. Stehle, C. Vira, D. Vira, D. Hogan, S. Feller, M. Affatigato, Phys. Chem. Glasses 39(2) (1998) 83. [12] A. Pan, A. Ghosh, J. Non-Cryst. Solids 271 (2000) 157-161. 17

ACCEPTED MANUSCRIPT

[13] P. Yasaka, N. Pattanaboonmee, H. J. Kim, P. Limkitjaroenporn, J. Kaewkhao, Ann. Nucl.

Energy 68 (2014) 4-9.

RI PT

[14] I. Kashif, A. A. El-Maboud, A. Ratep, Res. Phys. 4 (2014) 1-5. [15] Y. Dimitirev, A. C. Wright, V. Michajlova, E. Gattef, C. A. Guy, J. Mater. Sci Lett. 14 (1995) 347-350.

SC

[16] A. Khanna, A. Saini, B. Chen, F. Gonzalez, C. Pesquera, J. Non-Cryst. Solids 373 (2013)

M AN U

34-41.

[17] Y. Zhou, Y. Yang, F. Huang, J. Ren, S. Yuan, G Chen, J. Non-Cryst. Solids 386 (2014) 9094.

TE D

[18] G. Zhao, Y. Tian, H. Fan, J. Zhang, L. Hu, J. Mater. Sci. Technol. 29 (2013) 209-214. [19] Y. K. Sharma, S. S. L. Surana, R. K. Singh, Indian J. Pure & Appl. Phys. 46 (2008) 239244.

EP

[20] N. Chanthima, J. Kaewkhao, C. Kedkaew, W. chewpraditkul, A. Pokaipisit, P. Limsuwan,

AC C

Prog. Nucl. Sci. Tech. 1 (2011) 106-109. [21] J. M. Senior, Optical Fiber Communications: Principles and Practice, (UK) second ed., Prentice Hall International, UK, 2009, pp. 311-325. [22] X. Zhu, C. Mai, M. Li, J. Non-Cryst. Solids 388 (2014) 55–61. [23] K. Nanda, N. Berwal, R. S. Kundu, R. Punia, N. Kishore, J. Mol. Struct. 1088 (2015) 147154. 18

ACCEPTED MANUSCRIPT

[24] B. Shanmugavelu, V. V. Ravi Kanth Kumar, Solid State Sci. 20 (2013) 59-64. [25] Z. Pan, D. O. Henderson, S. H. Morgan, J. Non-Cryst. Solids 171 (1994) 134-140.

RI PT

[26] M. Abdel-Baki, F. A. Abdel-Wahab, F. El-Diasty, J. Appl. Phys. 111 (2012) 073506(1-6). [27] H. Doweidar, Y. B. Saddeek, J. Non-Cryst. Solids 355 (2009) 348-354.

SC

[28] W. Stambouli, H. Elhouichet, M. Ferid, J. Mol. Struct. 1028 (2012) 39-43.

M AN U

[29] O. Ravi, C. M. Reddy, L. Manoj, B. D. P. Raju, J. Mol. Struct. 1029 (2012) 53-59. [30] R. Kaur, S. Singh, K. Singh, O. P. Pandey, Radiat. Phys. Chem. 86 (2013) 23-30. [31] T. Kim, D. Gwoo, J. Kim, W. choi, K. Han, K. Kee, C. Hwang, B. Ryu, Electron. Mater.

Lett. 6 (2012) 617-620.

103 (2008) 1035099(0-6).

TE D

[32] B. Karthikeyan, C. S. S. Sandeep, J. Cha, H. Takabe, R. Philip, S. Mohan, J. Appl. Phys.

EP

[33] S. Rada, A. Dehelean, E. Culea, J. Non-Cryst. Solids 357 (2011) 3070-3073.

AC C

[34] G. Gao, L. Hu, H. Fan, G. Wang, K. Li, S. Feng, S. Fan, H. Chen, Opt. Mater. 32 (2009) 159-163.

[35] M. K. Halimah, W. M. Daud, H. A. A. Sidek, A. W. Zaidan, A. S. Zainal, Mater. Sci.

Poland 28 (2010) 173-180.

[36] E. A. Davis, N. F. Mott Philos. Mag. 22 (1970) 903-922. [37] J. Tauc, Mater. Res. Bull. 5 (1970) 721-730. 19

ACCEPTED MANUSCRIPT

[38] E. S. Moustafa, Y. B. Saddeek, E. R. Shaaban, J. Phys.Chem. Solids 69 (2008) 2281-2287. [39] P. L. Washington, H. C. Ong, J. Y. Dai, R. P. H. Chang, Appl. Phys. Lett. 72 (1998) 3261-

[40] F. Urbach, Phys. Rev. 92 (1953) 1324-1324.

SC

[41] V. Dimitrov, S. Sakka, J. Appl. Phys. 79 (1996) 1736-1740.

RI PT

3263.

[42] V. Dimitrov, T. Komatsu, J. Ceram. Soc. Jpn. 107 (1999) 1012-1018.

M AN U

[43] F. El-Diasty, F. A. Abdel Wahab, M. Abdel-Baki, J. Appl. Phys. 100 (2006) 093511-17.

AC C

EP

TE D

[44] Y. Wang, S. Dai, F. Chen, T. Xu, Q. Nie, Mater. Chem. Phys. 113 (2009) 407-411.

20

ACCEPTED MANUSCRIPT

Figure Captions Fig.1 X-ray diffraction patterns of different xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass

RI PT

compositions.

Fig.2 Composition dependence of molar volume (Vm) and crystalline volume (Vc) for xSiO2-(80-

x) Bi2O3-15B2O3-5TeO2 glass system.

SC

Fig.3 DSC curves for different xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass compositions.

Fig.4 Composition dependence of glass transition temperature (Tg) and oxygen packing density

M AN U

(OPD) for xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass system.

Fig. 5(a) FTIR spectra for different xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass compositions at

room temperature.

Fig. 5(b) Deconvoluted FTIR spectra for xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass system with

TE D

x=25.

Fig. 6(a) Raman spectra in the spectral range 100-850 cm-1; (b) Raman spectra in the spectral

temperature.

EP

range 850-1700 cm-1for different xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass compositions at room

Fig. 6(c) Deconvoluted Raman spectra in the spectral range 100-850 cm-1; (d) Deconvoluted

AC C

Raman spectra in the spectral range 850-1700 cm-1for xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass sample with x=20.

Fig. 7 Fitting of absorption coefficient (α) using Mott and Davis’s model (Eq. (5)) with r = 1/2,

2, 1/3 and 3 for xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass sample with x=15.

21

ACCEPTED MANUSCRIPT

Fig. 8 Fitting of absorption coefficient (α) using Mott and Davis’s model (Eq. (5)) with (a) r = 2

and (b) r = 3 for different compositions of xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass system. Fig. 9 Fitting of absorption coefficient (α) using Hydrogenic excitonic model (Eq. (6)) for

RI PT

different compositions of xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass system.

Fig. 10 Urbach’s plots for different compositions of xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass

AC C

EP

TE D

M AN U

SC

system.

22

ACCEPTED MANUSCRIPT

Table I. Density (ρ), molar volume (Vm), crystalline volume (Vc), optical packing density (OPD),

optical band gap# (Eg1), Urbach energy (∆E), molar refractivity (Rm) and metallization criterion

r=2

r=3

Eg1 (eV)

Eg1 (eV)

Vm

Vc

Tg

OPD

(gm/cm3)

(cm3/mole)

(cm3/mole)

(°C)

(g-atm/l)

0

5.995

65.06

46.93

472

45.34

2.451

5

5.892

62.75

45.45

475

46.22

2.422

10

5.878

59.45

43.96

478

47.94

15

5.827

56.49

42.48

484

20

5.811

53.15

40.99

472

25

5.734

50.33

39.51

471

30

5.697

47.09

38.03

#

469

∆E (eV)

Rm

αm

(cm3)

(Å3)

42.28

1.677

0.35

2.227

0.222

40.91

1.622

0.35

2.531

2.391

0.160

38.30

1.519

0.36

49.57

2.545

2.412

0.158

36.34

1.441

0.36

51.74

2.511

2.362

0.183

34.32

1.361

0.35

53.65

2.509

2.358

0.175

32.50

1.281

0.35

2.506

2.355

0.172

30.42

1.206

0.35

M AN U

0.206

56.28

AC C

EP

# parameters obtained from fitting of Mott-Davis model with experimental data of absorption coefficient (α).

M

2.272

TE D

ρ

SC

X

RI PT

(M) of xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glasses for different values of x.

23

ACCEPTED MANUSCRIPT

Table II. Peak position (xc), amplitude (A) and Full width at half maximum (W) of deconvoluted peaks of FTIR spectra of different

x=0

Num

xc -1

(cm )

A

W

(a.u.)

xc

x = 10 xc

A

W

(cm ) (cm )

(a.u.)

(cm )

(cm )

-1

-1

-1

-1

x = 15 xc

A

W

(a.u.)

(cm )

(cm )

-1

-1

x = 20 W

(a.u.)

(cm )

-1

474

7

92

467

10

90

469

4

64

459

5

82

2

556

4

96

546

5

97

529

7

115

527

6

114

3

708

3

77

711

3

69

708

4

80

707

2

77

4

831

5

55

832

4

53

834

7

53

831

2

5

876

18

83

876

15

82

878

13

75

873

6

968

38

122

967

40

131

969

52

146

973

7

1036

10

71

1051

16

95

1056

22

8

1090

9

68

1108

5

60

1116

5

9

1205

2.6

49

1214

3

48

1215

10

1243

5

69

1258

7

12

1311

13

119

1331

10

13

1522

6

96

1525

3

-1

(cm )

xc

A

W

(a.u.)

(cm )

(cm )

-1

-1

x=25

x=30

A

W

(a.u.)

(cm-1)

xc -1

(cm )

A

W

(a.u.)

(cm )

-1

472

14

110

476

15

105

449

6

75

552

4

`94

554

5

88

505

18

123

704

3

77

701

5

77

697

4

76

53

831

4

56

832

4

56

834

2

45

11

88

874

15

86

870

18

90

871

20

95

39

157

975

77

185

952

105

203

905

47

208

TE D

M AN U

1

110

1068

15

117

1064

17

123

1048

42

149

1025

115

209

69

1127

5

79

1124

7

86

1126

13

96

1136

11

101

52

1217

3

49

1217

4

52

1217

5

52

1220

4

47

EP 4

xc

A

AC C

ber

x=5

SC

Peak

RI PT

compositions of xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass system.

76

1261

10

85

1267

9

92

1270

20

107

1269

35

113

1269

35

110

104

1340

15

109

1354

11

102

1356

15

96

1356

25

101

1360

32

107

78

1533

6

92

1529

2

65

1538

4

78

1516

3

55

1505

1

32

24

ACCEPTED MANUSCRIPT

Table III. Peak position (xc), amplitude (A) and Full width at half maximum (W) of deconvoluted peaks of Raman spectra of different

Peak

x=0

x=5

x = 10

x = 15

RI PT

compositions of xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass system. x = 20

Number A

W

xc

A

W

xc

A

W

xc

A

W

xc

A

W

xc

(cm-1)

(a.u.)

(cm-1)

(cm-1)

(a.u.)

(cm-1)

(cm-1)

(a.u.)

(cm-1)

(cm-1)

(a.u.)

(cm-1)

(cm-1)

(a.u.)

(cm-1)

(cm-1)

1

113

10

91

113

10

91

113

9

91

114

9

91

xc

A

W

(cm-1)

(cm-1)

(a.u.)

(cm-1)

8

89

114

8

92

114

8

92

2

148

1

32

149

1

32

149

1

32

149

1

32

150

1

31

149

1

31

149

1

31

3

294

28

331

295

28

331

298

27

325

298

27

325

297

27

342

291

24

323

291

25

330

4

420

0.4

71

420

0.4

71

426

0.3

62

427

0.3

64

420

0.3

69

422

0.4

72

422

0.4

70

5

520

7

401

534

6

387

540

6

386

550

6

368

570

4

356

524

6

381

520

7

439

6

738

0.2

54

738

0.2

54

739

0.1

7

950

7

286

1000

10

397

1006

10

8

1208

66

184

1206

62

177

1203

9

1325

82

101

1325

86

103

1324

10

1409

34

95

1409

31

92

11

1488

31

113

1486

33

12

1613

179

313

1614

178

TE D

M AN U

114

A

x=30

W

SC

xc

x=25

(a.u.)

740

0.1

50

741

0.2

53

743

0.2

54

743

0.2

52

402

1010

10

343

1019

10

452

1105

14

582

1109

13

498

65

178

1172

35

151

1202

52

174

1200

47

169

1156

15

126

92

104

1317

39

87

1322

72

102

1322

73

104

1315

24

84

AC C

EP

50

1410

33

93

1372

164

246

1407

28

94

1406

22

88

1371

158

286

116

1488

30

111

1508

6

82

1487

22

108

1481

27

116

1504

5

80

310

1609

201

324

1627

171

297

1606

163

328

1611

150

317

1640

115

272

25

ACCEPTED MANUSCRIPT

Table IV. Values of optical energy band gap (Eg) , excitonic binding energy (R) , line widths for m = 1 state (Г1), line width of continuum (Гc)

RI PT

and absorption strength (Co), Best fit perameter (R2) and Reduced chi square (χ2) obtained from the fitting of experimental data of

x

Eg (eV)

R (eV)

Г1 (eV)

Гc (eV)

SC

absorption coefficient spectrum with Hydrogenic excitonic model (Eq.(6)) for xSiO2-(80-x) Bi2O3-15B2O3-5TeO2 glass system.

Co

R2

χ2

3.025

0.034

0.158

5

2.844

0.043

0.176

10

2.879

0.037

0.125

15

2.896

0.043

20

2.886

0.048

25

2.875

0.039

30

2.873

0.039

0.018

1074.49

0.999

0.045

0.075

127.82

0.993

0.182

0.052

144.89

0.996

0.074

TE D

0

M AN U

(eV1/2/cm)

0.044

173.22

0.995

0.139

0.146

0.052

134.03

0.995

0.118

0.143

0.058

131.79

0.997

0.045

0.139

0.054

160.40

0.993

0.213

AC C

EP

0.119

26

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

x=30

Intensity (A.U.)

x=25

x=20

x=15

x=10

x=5

x=0

20

30

40

50 2

(in degree)

60

70

80

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

65

65 V m

50

50

45

45

40

40

3

55

35

35 0

5

10

15 X (mol %)

20

25

30

c

(cm

55

/mole)

60

V c

V

V

m

(cm

3

/mole)

60

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

x=30 x=25

(W/g C)

x=20 x=15

dH/dT

x=10 x=5 T g

300

400

500

x=0

600

Temperature (

C)

700

800

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

485

56

T g OPD

54

475

50

T

g

( C)

52

48 470 46

465

44 0

5

10

15 X (mol %)

20

25

30

OPD (g-atm/l)

480

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

(a) x=30

(A.U.)

x=25

Absorbance

x=20

x=15

x=10

x=5

x=0

400

600

800

1000

Wavenumber

-1 (cm )

1200

1400

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

x=25

(b)

Fit Peak 1 Fit Peak 2 Fit Peak 3 Fit Peak 4

(A.U.)

Fit Peak 5 Fit Peak 6 Fit Peak 7

Absorbance

Fit Peak 8 Fit Peak 9 Fit Peak 10 Fit Peak 11 Fit Peak 12 Cumulative Fit Peak

400

600

800

1000

Wavenumber

1200 -1 (cm )

1400

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

(a) x=30

x=25

Raman Intensity

x=20

x=15

x=10

x=5

x=0

200

400

600 -1

Raman Shift (cm

)

800

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

(b) x=30

x=25

Raman Intensity

x=20

x=15

x=10

x=5

x=0

1000

1200 Raman Shift

1400 -1 (cm )

1600

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

(c)

Normalized Raman Intensity

x=20 Fit Peak 1 Fit Peak 2 Fit Peak 2 Fit Peak 3 Fit Peak 4 Fit Peak 5 Cumulative Fit Peak

100

200

300

400

500

600

-1 Raman Shift (cm )

700

800

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

(d) x=20

Normalized Raman Intensity

Fit Peak 1 Fit Peak 2 Fit Peak 3 Fit Peak 4 Fit Peak 5 Fit Peak 6 Cumulative Fit Peak

900

1000

1100

1200

1300

Raman Shift

1400 -1 (cm )

1500

1600

1700

40 x=15

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Theoretical Fit For r = 2

EP

Theoretical Fit For r = 3

Theoretical Fit For r = 1/2

30

AC C

-1

(cm )

Theoretical Fit For r = 1/3

20

10

0 2.6

2.7

2.8

h

(eV)

2.9

45

x=0

(a)

EP

x=5

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

x=10 x=15

AC C

30

-1

(cm )

x=20 x=25 x=30

Theoretical Fit

15

0 2.0

2.2

2.4

2.6

h

(eV)

2.8

3.0

x=0

40 x=5

(b)

EP

x=10

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

x=15 x=20

AC C

30

-1

(cm )

x=25 x=30

Theoretical Fit

20

10

0 2.0

2.2

2.4

2.6

h

(eV)

2.8

45

x=0 x=5

EP

x=10

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

x=15 x=20

AC C

30

-1

(cm )

x=25 x=30

Theoretical Fit

15

0 2.0

2.2

2.4

2.6

h

(eV)

2.8

3.0

4 x=0 x=5

EP

x=10

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

x=15

3

AC C

ln

(cm

-1

)

x=20 x=25 x=30

Linear Fit

2

1

2.0

2.2

2.4

2.6

h

eV)

2.8

3.0

ACCEPTED MANUSCRIPT Highlights: SiO2 is present in SiO4 tetrahedral structural units with two NBOs per silicon.



Bi2O3 is present in both BiO6 octahedral and BiO3 pyramidal structural units.



HEM shows good agreement with the experimental observed absorption spectra.

AC C

EP

TE D

M AN U

SC

RI PT