Upload
lamngoc
View
223
Download
1
Embed Size (px)
Citation preview
Optical constants of silica glass from extreme ultravioletto far infrared at near room temperature
Rei Kitamura,1 Laurent Pilon,1,* and Miroslaw Jonasz2
1Department of Mechanical and Aerospace Engineering, Henry Samueli School of Engineering and Applied Science,University of California, Los Angeles, Los Angeles, California 90095, USA
2MJC Optical Technology, 217 Cadillac Street, Beaconsfield, QC H9W 2W7, Canada
*Corresponding author: [email protected]
Received 30 July 2007; accepted 31 August 2007;posted 25 September 2007 (Doc. ID 85883); published 19 November 2007
We thoroughly and critically review studies reporting the real (refractive index) and imaginary (absorp-tion index) parts of the complex refractive index of silica glass over the spectral range from 30 nm to1000 �m. The general features of the optical constants over the electromagnetic spectrum are relativelyconsistent throughout the literature. In particular, silica glass is effectively opaque for wavelengthsshorter than 200 nm and larger than 3.5–4.0 �m. Strong absorption bands are observed (i) below 160 nmdue to the interaction with electrons, absorption by impurities, and the presence of OH groups and pointdefects; (ii) at �2.73–2.85, 3.5, and 4.3 �m also caused by OH groups; and (iii) at �9–9.5, 12.5, and21–23 �m due to SiOOOSi resonance modes of vibration. However, the actual values of the refractiveand absorption indices can vary significantly due to the glass manufacturing process, crystallinity,wavelength, and temperature and to the presence of impurities, point defects, inclusions, and bubbles,as well as to the experimental uncertainties and approximations in the retrieval methods. Moreover,new formulas providing comprehensive approximations of the optical properties of silica glass areproposed between 7 and 50 �m. These formulas are consistent with experimental data and substantiallyextend the spectral range of 0.21–7 �m covered by existing formulas and can be used in various engi-neering applications. © 2007 Optical Society of America
OCIS codes: 160.6030, 160.4760, 290.3030, 160.4670, 050.5298, 110.5220.
1. Introduction
Silicon dioxide (SiO2 or silica) has many forms includ-ing three main crystalline varieties: quartz, tridym-ite, and cristobalite [1,2]. Silica can also exist innoncrystalline form as silica glass or vitreous silica[1], also referred to as amorphous silica and glassysilica. There are four basic types of commercial silicaglasses [3,4]: (1) Type I is obtained by the electricmelting of natural quartz crystal in vacuum or in aninert gas at low pressure, (2) Type II is produced fromquartz crystal powder by flame fusion, (3) Type III issynthetic and produced by the hydrolyzation of SiCl4when sprayed into an OH flame, and (4) Type IV isalso synthetic and produced from SiCl4 in a watervapor-free plasma flame.
Each type of silica glass has its own impurity leveland optical properties. For example, Type I silicaglasses tend to contain metallic impurities [3,4]. Onthe other hand, Types III and IV are much purer thanType I and feature greater ultraviolet (UV) transmis-sion [3,4]. However, Type III silica glasses have, ingeneral, higher water content, and infrared (IR)transmission is limited by strong water absorptionpeaks at wavelengths around 2.2 and 2.7 �m [3–5].Type IV is similar to Type III but contains less waterand thus has better IR transmission. Suppliers pro-vide various product grades for different optical ap-plications [3,4,6]. Trade names for Type I silica glassare Infrasil, IR-Vitreosil, Pursil, and GE 105, for ex-ample. On the other hand, KU, Herosil, Homosil,Vitrasil, and OG Vitreosil are Type II. Moreover,KU-1, KV, Suprasil I, Tetrasil, Spectrosil, and Corn-ing 7940 are known as synthetic fused silica andclassified as Type III. Finally, KI, Suprasil W, Spec-
0003-6935/07/338118-16$15.00/0© 2007 Optical Society of America
20 November 2007 � Vol. 46, No. 33 � APPLIED OPTICS 8118
trosil WF, and Corning 7943 are Type IV silica glass.The term “fused quartz” is usually used for Type Iand “fused silica” is used for synthetic fused silica,i.e., Types III or IV. In practice, however, those termsare often used interchangeably.
Because of its favorable physical, chemical, andoptical characteristics, silica glass has been used innumerous applications: (1) as laboratory glassware,(2) in optics, as lenses or beam splitters, (3) for light-ing and IR heating, (4) in telecommunications, asfiber optics, (5) in micro- and optoelectronics, as adielectric insulator, a waveguide, photonic crystal fi-bers, or projection masks for photolithography, and(6) in thermal protection systems, as fibrous thermalinsulation. In all these applications, optical proper-ties are essential to predict and optimize the opticaland thermal radiation performances of this material.Silica fiber optics, for example, are used in thenear-IR at �1.31 and 1.55 �m due to their low opticalattenuation and optical dispersion [7,8]. In lens de-signing, one often needs to fit and interpolate refrac-tive index data that are reported or measured atdiscrete wavelengths over certain spectral regions[9,10]. On the other hand, astronomers and atmo-spheric scientists are interested in the optical prop-erties of interstellar and atmospheric silica particlesin the mid- and far-IR region of the spectrum [11–13].
The complex refractive index, m�, of silica glass atwavelength � is defined as
m� � n� � ik�, (1)
where n� is the refractive index and k� is the absorp-tion index. The wavelength � is related to otherquantities such as frequency � and wavenumber �according to
� �c�
��
1�
, (2)
where c� is the speed of light at wavelength � invacuum. Therefore, the refractive and absorption in-dices can also be expressed as functions of frequencyand denoted by n� and k� or as functions of wavenum-ber and denoted by n� and k�.
The experimental data for the refractive and ab-sorption indices vary in precision depending on themeasurement techniques used and on the approxi-mations made in retrieving the intrinsic optical prop-erties. One should also keep in mind that theseoptical properties may be sensitive to the presence ofimpurities, crystallinity, point defects, inclusions,bubbles, wavelength, temperature, and the glassmanufacturing process. In addition, when consider-ing the literature it is often difficult to choose the setof experimental data or formula to use and to assessits validity. The objectives of this paper are (1) tocritically review and compare the experimental datareported in the literature for the complex refractiveindex of silica glass and (2) to develop formulas thatprovide a comprehensive approximation of the mea-
sured data near room temperature. Given the widerange of engineering and scientific applications, thespectral range from 30 nm to 1000 �m is considered.
2. Experimental Methods
A. Refractive Index, n�
Various experimental techniques and procedureshave been used to retrieve the real part of the com-plex refractive index n�. The most accurate is theminimum deviation angle method [14], which relieson measuring the minimum deviation angle �min of anisosceles triangular prism made of the silica glassplaced in air. This method is based on Snell’s law [14],and the refractive index n� can be estimated by
n� �
sin��min � �
2 �sin��
2�nair, (3)
where � is an apex angle of the prism sample and nairis the refractive index of air �nair � 1�. This method isoften used to accurately measure the refractive indexof highly transparent glass for which the absorptionindex k�, i.e., the imaginary part of the complex re-fractive index, is negligibly small.
Alternatively, the interferometric method is alsoused to measure n�. It is based on observing the in-terference fringes created when light is incidentnormally upon a silica glass plate [15,16]. Other tech-niques include the Abbe’s or the Pulfrich’s refractom-eters whose accuracy on the index of refraction iswithin 2 10�3 and 5 10�5, respectively [17].
Moreover, when absorption cannot be ignored, bothn� and k� can be retrieved from the directional orhemispherical reflectance and�or emittance of a slabof known thickness. Electromagnetic wave theory canbe used to retrieve both n� and k�, assuming opticallysmooth surfaces and accounting for internal reflec-tion [18]. Finally, the Kramers–Krönig relations [18]can also be used to predict either the refractive indexfrom the absorption index, or vice versa, at frequency� [18,19]:
n� � 1 �2�
P�0
��k��
��2 � �2 d��, (4)
k� ��2�
�P�
0
n��
��2 � �2 d��, (5)
where P denotes the Cauchy principle value of theintegral [18]. Alternatively, the refractive and ab-sorption indices can be simultaneously obtained fromreflectance. First, the phase angle of the complex re-flection coefficient �, at frequency �0 can be expressedby the Kramers–Krönig relations [20]:
8119 APPLIED OPTICS � Vol. 46, No. 33 � 20 November 2007
���� �2�
�P�
0
d ln� R������2 � �2 d��, (6)
where R��� is the normal–normal reflectivity ex-pressed as a function of �:
R��� � �r�2 ��n� � 1�2 � k�
2
�n� � 1�2 � k�2. (7)
The Fresnel reflection coefficient r is defined as
r �1 � n� � ik�
1 � n� � ik�� �r�e
i�. (8)
From Eqs. (7) and (8), the refractive and absorptionindices can be expressed as
n� �1 � R���
1 � R��� � 2�R��� cos �, (9)
k� ��2�R��� sin �
1 � R��� � 2�R��� cos �. (10)
However, due to the infinite bound of the integrals inEqs. (4)–(6), these techniques require extrapolationsinto spectral regions where data are not always avail-able. Furthermore, the integrals need to be computednumerically. Practical limitations and possible errorsof Kramers–Krönig relations have been discussed byRiu and Lapaz [19]. These authors concluded that theKramers–Krönig relations were practically applica-ble in almost every experimental situation.
B. Absorption Index, k�
The value of the absorption index k� was not alwaysdirectly available from the literature and was some-times recovered from the normal spectral transmit-tance or emittance data. Indeed, the value of k� can berecovered from the normal spectral transmittanceT0,�, accounting for multiple reflections and expressedas [21]
T0,��L� ��1 � ���2e���L
1 � ����2e�2��L, (11)
where L is the thickness of the sample and �� and ��
are the spectral reflectivity of the interface and thespectral absorption coefficient of silica glass, respec-tively, and are given by
�� ��n� � 1�2 � k�
2
�n� � 1�2 � k�2, (12)
�� �4�k�
�. (13)
Equations (11)–(13) can be solved as a quadratic inthe exponential factor in terms of k�. After some al-gebraic manipulation, one obtains the following ex-pression for k� as a function of the refractive index, n�,the sample thickness, L, and the spectral normaltransmittance, T0,�:
k� � �� �
4�L�ln��1 � ���4 � 4��2T0,� � �1 � ���
2��2T0,�
.
(14)
Alternatively, the absorption index, k�, can also bedetermined from measurements of the spectral nor-mal emittance, ��,0, by using the following expression[22]:
k� � � �
4�L�ln1 � �� � ����,0
1 � �� � ��,0. (15)
Note that the above expressions for the absorptionindex k� given by Eqs. (14) and (15) are valid if both�n� � 1� and �n� � 1� are much larger than k�. Ineither case, an expression for n� is necessary to esti-mate ��. As discussed later in this paper, theSellmeier equation proposed by Malitson [23] can beused for that purpose between 0.21 and 6.7 �m.
In addition, in the UV and IR regions of the spec-trum when silica glass is strongly absorbing, mostreported values of n� and k� were retrieved from near-normal reflectance measurements in combinationwith the Kramers–Krönig relations [24].
3. Experimental Data and Discussion
Table 1 summarizes representative references re-porting experimental values of the complex refractiveindex of silica glass at room temperature for the spec-tral range from 30 nm to 1000 �m. For each study,the measurement method, the spectral range, as wellas the sample thicknesses, compositions, and temper-atures investigated are also reported when available.In addition, the absorption index k� was derived fromtransmittance measurements using Eq. (14) if it wasnot directly reported. Then, computation from trans-mittance and emittance data sometimes leads to neg-ative values, particularly in the spectral region from0.2 to 4.0 �m where silica glass is very weakly ab-sorbing. This was the case for transmittance datafrom [25,26]. Hence, in this region, the experimentaldata should be used with care since the uncertaintyfor k� is very large and k� effectively vanishes.
Figure 1 shows the real and imaginary parts of therefractive index of silica glass, n� and k�, as a functionof wavelength � over the spectral range from 30 nm to1000 �m as reported in the references listed in Table1. Because of the density of data points in some partsof the spectrum and for the sake of clarity, Figs. 2–5show details of both the real n� and imaginary k�
parts of the complex index of refraction of silica glassfor wavelengths between 30 nm and 1 �m, 1 and
20 November 2007 � Vol. 46, No. 33 � APPLIED OPTICS 8120
Tab
le1.
Sum
mar
yo
fth
eE
xper
imen
talD
ata
Rep
ort
ing
the
Co
mp
lex
Ind
exo
fR
efra
ctio
no
fS
ilica
Gla
ssat
Ro
om
Tem
per
atur
ea
Ref
eren
ceW
avel
engt
hR
ange
(�m
)M
easu
rem
ent
Met
hod
Rep
orte
dD
ata
Tem
pera
ture
Sam
ple
Thi
ckne
ssT
ype
Com
men
ts
[11]
7.14
–25
Refl
ecti
on,t
rans
mis
sion
,K
ram
ers–
Krö
nig
rela
tion
n,k
RT
N�A
IV
itre
osil
[12]
2–50
0R
eflec
tion
,Kra
mer
s–K
röni
gm
etho
dR
,n,k
300,
200,
100,
and
10K
N�A
III
Sup
rasi
l,U
Vgr
ade
synt
heti
cfu
sed
SiO
2
[13]
7–30
0T
rans
mis
sion
n,k
RT
N�A
N�A
[15]
110–
550
Inte
rfer
omet
ric
T,n
,kR
T2.
1409
mm
IIn
fras
il(l
owH
2O
)[1
6]3–
6.7
Inte
rfer
omet
ric
n,T
RT
0.23
mm
III
Sup
rasi
l2
�10
00pp
mO
Hco
nten
t[2
3]0.
21–3
.71
Min
imum
devi
atio
nm
etho
dn
20°C
N�A
III
Cor
ning
code
7940
,Dyn
asil
high
-pur
ity
synt
heti
cS
iO2
glas
san
dG
Ety
pe15
1[2
5]0.
80–2
.60
kfr
omE
q.(1
4)T
298
K1.
6m
mN
�A[2
6]0.
19–0
.42
kfr
omE
q.(1
4)T
298
K2
mm
N�A
[27]
0.05
–0.7
Refl
ecti
onan
dK
ram
ers–
Krö
nig
n,�
RT
N�A
III
Sup
rasi
l(U
Vgr
ade)
[28]
7.19
–9.0
6In
terf
erom
etri
cn,
T,R
RT
0.19
56m
mIV
Sup
rasi
lW
2[2
9]0.
2–3
Refl
ecti
onan
dtr
ansm
issi
onn,
k,R
,T24
°CN
�AN
�A[3
0]16
.7–2
5R
eflec
tion
R,n
,kR
TN
�AN
�A[4
4]0.
35–3
.51
Min
imum
devi
atio
nm
etho
dn
24°C
N�A
N�A
Sam
ples
from
Gen
eral
Ele
ctri
c,H
erae
us,N
iede
rF
used
Qua
rtz,
and
Cor
ning
Gla
ssW
orks
[45]
1.44
–4.7
7In
terf
erom
etri
cn,
T23
.5°C
–481
°C0.
1994
mm
IVS
upra
sil
W2
[49]
7.14
–50
Refl
ecti
onn,
kR
TN
�AII
KU
and
KI
Kra
mer
s–K
röni
gre
lati
onIV
Impu
rity
did
not
exce
ed0.
007%
[50]
0.00
06–5
00n,
kR
TN
�AN
�AC
ompi
lati
onof
data
[51]
1.00
–7.5
Tra
nsm
issi
onT
,R
T5
�m
to3.
06m
N�A
[54]
0.10
–0.1
6R
eflec
tion
and
Kra
mer
s–K
röni
gn,
kR
TN
�AN
�AS
ampl
espr
oduc
edby
Ele
ctro
-Qua
rtz
[55]
1.35
–4.8
5In
terf
erom
etri
cn,
T23
.5°C
0.23
45m
mIV
Sup
rasi
lW
2[5
6]8.
13–9
.63
Refl
ecti
onR
,n,k
RT
N�A
[57]
7.84
–12.
9R
eflec
tion
n,k
RT
N�A
[58]
0.23
–3.3
7M
inim
umde
viat
ion
met
hod
n26
°C,4
71°C
,and
828
°CN
�AII
IC
orni
ng79
40
[59]
7.14
–11.
11R
eflec
tion
,tra
nsm
issi
on,a
ndK
ram
ers–
Krö
nig
rela
tion
n,k
RT
N�A
IIK
UII
IK
VIV
KI
IIn
fras
il[6
0]7.
14–1
4.29
Refl
ecti
onR
,kR
TN
�AN
�A[6
1]0.
5–4.
5M
inim
umde
viat
ion
nR
TN
�AN
�AS
ampl
eha
ving
low
wat
erco
nten
tsu
ppli
edby
Gen
eral
Ele
ctri
cC
ompa
ny[6
2]0.
37,0
.44,
0.55
,1.
01,a
nd1.
53M
inim
umde
viat
ion
n29
4,24
0,18
0,an
d12
0K
N�A
N�A
[63]
0.06
–40
n,k
RT
N�A
N�A
[64]
0.2–
3.5
Tra
nsm
issi
onk,
TR
T—
1500
°C0.
953
mm
I,II
IG
EIR
type
105,
GE
UV
type
151,
and
Cor
ning
Vyc
orIR
Gla
ssN
umbe
r79
05[6
5]2.
00–6
.00
kfr
omE
q.(1
4)T
25°C
and
400
°C2.
8m
mI
GE
Typ
e10
5[6
6]0.
31–3
.97
kfr
omE
q.(1
4)T
RT
5.45
mm
N�A
8121 APPLIED OPTICS � Vol. 46, No. 33 � 20 November 2007
15 �m, 15 and 100 �m, and 100 and 1000 �m, re-spectively.
Overall, the reported values for both the real andimaginary parts of the complex index of refractionagree relatively well. Studies showing large devia-tions from other studies suggest that the data areunreliable [24]. For example, data reported by Elliset al. [27] strongly disagree with all other data be-tween 60 nm and the visible. Locations of extrema ofn� are consistent among all experimental data ex-cept for those reported by Tan [28] �7.19 �m � �� 9.06 �m�, Khashan and Nassif [29] �0.2 �m � �� 3.0 �m�, and Reitzel [30] �16.7 �m � � � 25 �m�.The data for wavelengths below 9 �m as reported byvarious authors agree well with one another. How-ever, the data agree considerably less for wave-lengths between 9 and 50 �m. Beyond 50 �m, asmaller number of values for n� have been reported,but the data agree relatively well.
Furthermore, trends and the locations of extremain the measured absorption index are consistent fromone study to another. However, discrepancies largerthan those for n� can be observed in reported data fork� in some parts of the spectrum. They are most likelydue to (i) the impurity of the sample (e.g., OH group,alkali, metallic content), (ii) the presence of inclu-sions, bubbles, or point defects, (iii) the sample prep-aration and surface optical quality, and�or (iv) theuncertainty in the measurement and retrieval tech-niques. Note that the flatness of the sample surfacebecomes a critical parameter in the visible and UV[31,32]. In this wavelength range, the surface rough-ness must remain much smaller than the wavelengthto avoid surface scattering and consider the surfaceas optically smooth.
The imaginary part of the complex refractive indexof silica glass k� is small from the near UV to thenear-IR part of the spectrum. Practically, silica glassis transparent from 200 nm up to 3.5–4.0 �m. In theextreme UV (for wavelengths below 200 nm) and inthe IR and far-IR (beyond 4.0) silica glass can beconsidered opaque. In the UV region of the spectrumbelow 200 nm, the strong absorption of silica glass iscaused by the interaction of the electromagnetic ra-diation with electrons of SiOO bonds [17] and withstructural imperfections or point defects such as OHgroups, SiOSi bonds, and strained SiOOOSi bonds[33]. This results in a sharp UV cutoff (also called theabsorption edge) at �160 nm [17,34]. The location ofthe absorption edge depends on the glass composi-tion, impurity level, and point defects formed duringthe manufacturing process [17,33] as well as on tem-perature [35]. For example, it is shifted toward thevisible wavelengths due to the presence of impuritiesin particular ions Fe3�, Cr3�, and Ti3� [36]. Similareffects are observed when increasing the alkali con-tents [34] or the temperature [35]. On the contrary,the absorption edge is slightly shifted to lower wave-lengths for crystal quartz [34]. Shifting the absorp-tion edge to lower wavelengths (even slightly) hasbeen the subject of intense studies to enable the use
[67]
0.16
–0.3
0k
from
Eq.
(14)
TR
T2.
04–3
.29
mm
III
Cor
ning
code
7940
and
Dyn
asil
[68]
1.00
–4.6
2k
from
Eq.
(14)
TR
T3.
18m
m[6
9]7.
69–1
1.11
Refl
ecti
onan
dK
ram
ers–
Krö
nig
n,k
RT
1.0
mm
IIK
UII
IO
ptic
algr
ade
fuse
dqu
artz
IVK
I[7
0]7.
41–5
0R
eflec
tion
and
disp
ersi
onan
alys
isb
n,k
RT
N�A
IVK
I[7
1]60
–560
Tra
nsm
issi
onan
dre
flect
ion
R,n
,kR
T0.
258,
1.05
,2.0
3,4.
07,1
2.35
,12.
35,
and
25m
m
N�A
Al 2
O3
�4.
0�
10�
3%
,Fe 2
O3
�4.
0�
10�
3%
,C
aO�
2.0
�10
�3%
,Na 2
O�
1.0
�10
�3%
,C
uO,T
iO2,M
gO,M
n 3O
4
0.00
1%[7
2]83
.3–5
00T
rans
mis
sion
n,
300
K1.
340
�0.
001
mm
Opt
ical
lypo
lish
ed[7
3]50
–100
0R
eflec
tion
and
tran
smis
sion
n,k
RT
N�A
N�A
Impu
rity
�3
�10
�5
[74]
100–
1000
Tra
nsm
issi
on
RT
N�A
GE
type
s10
1an
d10
6[7
5]0.
029–
1.77
Tra
nsm
issi
on,r
eflec
tion
,and
Kra
mer
s–K
röni
gT
,R,n
,kR
TN
�AII
IS
upra
sil
I
[76]
100–
667
Tra
nsm
issi
on
100
and
300
KG
Ety
pe10
1[7
7]7.
69–1
1.1
Inte
rnal
refle
ctio
nan
dK
ram
ers–
Krö
nig
n,k
RT
N�A
III
KU
-1gl
ass,
brok
ensu
rfac
e
[78]
2–35
Refl
ecti
onan
dK
ram
ers–
Krö
nig
T,R
,n,k
RT
N�A
N�A
aR
Tis
room
tem
pera
ture
.bR
efer
ence
[24]
.
20 November 2007 � Vol. 46, No. 33 � APPLIED OPTICS 8122
Fig. 1. Real n� and imaginary k� parts of the complex refractive index of silica glass reported in the literature and summarized in Table1. The solid curve (present study) was obtained with Eqs. (21)–(24) by using coefficients listed in Table 2.
8123 APPLIED OPTICS � Vol. 46, No. 33 � 20 November 2007
Fig. 2. Real n� and imaginary k� parts of the complex refractive index of silica glass between 30 nm and 1 �m as reported in the literatureand summarized in Table 1.
20 November 2007 � Vol. 46, No. 33 � APPLIED OPTICS 8124
Fig. 3. Real n� and imaginary k� parts of the complex refractive index of silica glass between 1 and 15 �m as reported in the literatureand summarized in Table 1. The solid curve (present study) was obtained with Eqs. (21)–(24) by using coefficients listed in Table 2.
8125 APPLIED OPTICS � Vol. 46, No. 33 � 20 November 2007
Fig. 4. Real n� and imaginary k� parts of the complex refractive index of silica glass between 15 and 100 �m as reported in the literatureand summarized in Table 1. The solid curve (present study) was obtained with Eqs. (21)–(24) by using coefficients listed in Table 2.
20 November 2007 � Vol. 46, No. 33 � APPLIED OPTICS 8126
Fig. 5. Real n� and imaginary k� parts of the complex refractive index of silica glass between 100 and 1000 �m as reported in the literatureand summarized in Table 1.
8127 APPLIED OPTICS � Vol. 46, No. 33 � 20 November 2007
of silica glass for photomask material in 157 nmphotolithography using F2 excimer lasers [32,37,38].An acceptable transmittance at �157 nm has beenachieved by minimizing the OH content of silica [37]or by doping silica glass with network modifiers suchas fluorine, which relaxes the glass structure andeliminates strained SiOOOSi bonds [33]. Experi-mental measurements and theoretical calculations ofthe electronic structure of SiO2 has been reviewed byGriscom [39] and spectroscopic data for wavelengthsbetween 90 and 350 nm have been discussed by Sigel[34]. An interaction between the UV radiation andelectrons and point defects is also responsible forthe steep increase of n� for wavelengths less than300 nm.
In the IR part of the spectrum, silica glass iseffectively opaque for a wavelength larger than3.5–4.0 �m. Beyond this wavelength, three major ab-sorption bands can be observed (Fig. 1) due to theresonance of SiOOOSi vibrations. The absorptionpeak between 9.0 and 9.5 �m can be attributed tothe asymmetric stretching vibration of SiOOOSibridges [17,24]. The absorption band at �12.5 �m isdue to the symmetric vibration stretching of theSiOOOSi bridge involving the displacement of the Oatom perpendicular to the SiOSi direction in theSiOOOSi plane [24]. The third band between 21 and23 �m is the consequence of the OOSiOO bendingvibration but has also been attributed to the “rocking”mode of SiOOOSi bonds caused by the displacementof an oxygen atom out of the SiOOOSi plane [24].The resonance of SiOOOSi vibrations are also re-sponsible for the sharp decreases in n� around theresonance wavelengths [17]. The reader is referredto [24, pp. 63–77] for a detailed discussion on vi-brational spectroscopy of silica glass at these wave-lengths. Moreover, smaller absorption bands at� wavelengths 2.73–2.85, 3.5, and 4.3 �m correspondto the presence of OH groups in the structure of theglass [17,40–42]. The magnitude of the absorptiondepends on the melting technology and in particularon the partial pressure of water vapor above the meltduring the melting process [17]. The concentration ofOH groups in silica glass can be computed from theabsorption band at � wavelength 2.73–2.85 �m [17].To the best of our knowledge, no model or approxi-mate equation has been proposed for the absorptionindex of silica glass. This is the subject of Section 4.
4. Optical Constant Theory
The complex index of refraction, m� � n� � ik�,and the complex relative dielectric permittivity,���� � ����� � i����� are related by the expression���� � m�
2, i.e., [21]
����� � n�2 � k�
2, � ���� � 2n�k�. (16)
Numerous physical models such as the Lorentzmodel, the Drude model, and the Debye relaxationmodel have been proposed to predict the optical prop-erties of solids [18]. The Lorentz model assumes
that electrons and ions in the material are harmonicoscillators subject to the force applied by time-dependent electromagnetic fields. Then the complexrelative dielectric permittivity can be expressed interms of frequency � as follows [21]:
���� � 1 � �j
�pj2
�j2 � �2 � i�j�
� 1 � �j
�pj2��j
2 � �2� � i�j�pj2�
��j2 � �2�2 � �j
2�2 , (17)
where �pj and �j are the plasma and resonance fre-quencies, respectively. The parameter �j is the damp-ing factor of the oscillators. Only when � is very closeto one of the resonance frequencies �j, the imaginaryterms in Eq. (17) are important [43]. Thus, �j� arenegligibly small compared with ��j
2 � �2� for silicaglass for a wavelength below 7 �m and �� is virtuallyequal to 0.0. Hence, after substituting Eq. (2) intoEq. (17), � can be expressed in terms of � as follows:
���� � ����� � 1 � �j
Aj2�2
��2 � �j2�
, (18)
where Aj � �pj�j�c� with �j being the resonance wave-length. Moreover, as ����� vanishes, the medium isweakly absorbing and k� is negligibly small comparedwith n�. Then ����� is equal to n�
2 and given by theSellmeier dispersion formula:
����� � n�2 � 1 � �
j
Aj2�2
��2 � �j2�
. (19)
Different formulas for the refractive index of silicaglass as a function of wavelength and based on theSellmeier dispersion formula have been proposed inthe literature [23,44,45] for different spectral regions.Rodney and Spindler [44] suggested a formula for n�
over a spectral range from 0.347 to 3.508 �m at 31 °Cwhile Tan and Arndt [45] proposed another equationin the spectral region from 1.44 to 4.77 �m at tem-peratures ranging from 23.5 to 481 °C. In addition,for the spectral range from 0.21 to 3.71 �m at 20 °C,Malitson [23] fitted experimental data with the fol-lowing three-term Sellmeier equation:
n�2 � 1 �
0.6961663�2
�2 � �0.0684043�2�
0.4079426�2
�2 � �0.1162414�2
�0.8974794�2
�2 � �9.896161�2. (20)
Tan [16] confirmed the validity of Eq. (20) for wave-lengths up to 6.7 �m. Furthermore, for a spectralrange over 8 �m, an approximate piecewise linear fitwas given by Dombrovsky [46]. However, no physics-
20 November 2007 � Vol. 46, No. 33 � APPLIED OPTICS 8128
based formulas have been developed for the spectralrange beyond 8 �m.
In parts of the spectrum where k� cannot be ne-glected or when the frequency � is very close to theresonance frequencies, the Sellmeier equation for n�
is no longer valid and an alternative model must beused. Recently, Meneses et al. [47] proposed a newdielectric function model based on the causal versionof the Voigt function. The model was validated byfitting the IR spectra of two different glasses and
confirmed to be more appropriate than the Lorentzmodel [47]. Moreover, the authors proposed anothersimplified model based on Gaussian functions [48].Then, the dielectric constant can be written as
���� � ����� � i� ���� � � � �j
�gcjkkg��� � igcj��� ,
(21)
where the high frequency dielectric constant is de-noted by � . In addition, the Gaussian functions gcj���and gcj
kkg��� are defined as
gcj��� � �j exp�4 ln 2�� � �0j
�j�2� �j
exp�4 ln 2�� � �0j
�j�2, (22)
gcjkkg��� �
2�j
� D�2�ln 2� � �0j
�j�� D�2�ln 2
� � �0j
�j�.
(23)
Here, �j is the amplitude, �0j is the peak position, �j isthe full width at half-maximum, and D�x� is an oper-ator defined as
Fig. 6. Residuals between experimental [49] and predicted values of n� and k�. The predicted values are based on Eqs. (21)–(24) withcoefficients listed in Table 2.
Table 2. Parameters Used to Interpolate the Refractive Index n� andAbsorption Index k� of Silica Glass by Using Eqs. (21)–(24)a,b
j �j �0j �j
1 3.7998 1089.7 31.4542 0.46089 1187.7 100.463 1.2520 797.78 91.6014 7.8147 1058.2 63.1535 1.0313 446.13 275.1116 5.3757 443.00 45.2207 6.3305 465.80 22.6808 1.2948 1026.7 232.14
a�� � 2.1232.bThese parameters were obtained by fitting the equations to data
of Popova et al. [49].
8129 APPLIED OPTICS � Vol. 46, No. 33 � 20 November 2007
Fig. 7. Residuals between the experimental data on the refractive index and absorption index and values predicted in this work by usingEqs. (21)–(24) along with coefficients listed in Table 2.
20 November 2007 � Vol. 46, No. 33 � APPLIED OPTICS 8130
D�x� � e�x2�0
x
et2
dt. (24)
In the present study, this model is used to interpolatethe refractive index n� and absorption index k� forwavelengths � between 7 and 50 �m. It enables one todescribe the experimental data with a reduced set ofparameters [48] over a wide spectral range, includingthe spectral range where k� may be large. Note that theabove simplified model satisfies the Kramers–Krönigrelation.
The practical procedure for fitting complex refrac-tive index data conducted in this paper is as follows:(i) the spectral reflectivity at normal incidence, R���,is computed from Eq. (7), (ii) parameters of ���� inEqs. (21) and (22) are determined by curve fitting forR��� expressed as
R��� � ������ � 1
����� � 1�2
, (25)
and (iii) n� and k� are computed using Eq. (16). Theadvantage of this procedure is that (a) fitting thereflectivity is easier than fitting n� and k� indepen-dently, (b) both n� and k� can be derived from a singlecurve fitting, and (c) the result automatically satisfiesthe Kramers–Krönig relations.
The experimental data of Popova et al. [49] wereselected to develop formulas for both n� and k� be-cause these data cover a wide spectral range from 7 to50 �m, and both the refractive and absorption indicesare reported at the same wavelength enabling thecalculation of R���. In the spectral region from 0.2 to7 �m, the absorption index of silica glass is verysmall and may be assumed to be zero for all practicalpurposes as suggested by Fig. 1. Moreover, reporteddata, including that of Popova et al. [49], indicatesthat the refractive index is satisfactorily predicted bythe Sellmeier formula reported by [16] and Malitson[23] between 0.2 and 7 �m and given by Eq. (20).Therefore, the present study focuses on the spectralrange between 7 and 50 �m.
To fit the model with experimental data, the FOCUS
software was used [47,48]. By adding terms in Eq.(21) one by one, eight terms were found to best fit theexperimental data. The fitting curves obtained in thisstudy are shown in Figs. 3 and 4, and the associatedparameters �j, �0j, and �j used in Eqs. (21)–(23) aresummarized in Table 2. The fitting curves for n� andk� obtained in the present study agree well with thedata of Popova et al. [49] and with most of other datashown in Figs. 3 and 4. The differences (or residuals)between the data of Popova et al. [49] and the modelpredictions for both n� and k� using the parametersgiven in Table 2 are shown in Fig. 6. It indicates thatthe residuals are less than 0.06 except near 9 and22.5 �m, where they reach up to 0.3 around 9 �m and0.14 around 22.5 �m. This can be attributed to the
fact that the refractive index changes greatly in thosewavelength regions. However, these residuals aresmall compared with much larger difference observedin the experimental data reported for n�. For exam-ple, one can see differences larger than 0.5 amongreported refractive index data in the spectral rangebetween 9 and 10 �m. Thus, the approximation ob-tained here is acceptable and can be useful in engi-neering applications.
Moreover, the residuals between the experimentaldata for the refractive and absorption indices listed inTable 1, and the model predictions using the param-eters given in Table 2 are shown in Fig. 7. Here again,the residuals are less than 0.3 except at �9.0 and22.5 �m. This indicates that the model predictionsusing the parameters given in Table 2 agree with alldata reported in the literature at the spectral rangeexcept at �9 and 22.5 �m. However, large discrep-ancies in reported experimental data sets can be ob-served around these two wavelengths. In general, themodel should be used with care when applied outsidethe spectral range from 0.2 to 50 �m.
Finally, given its widespread use, particular atten-tion was paid to the compilation of data reported byPhilipp [50] for the refractive and absorption indicesof silica glass over the spectral range from 0.006 to500 �m. The compilation consists of experimentaldata reported by various authors [15,20,51–54] aswell as unpublished data. Philipp [50] also retrievedoptical properties from the absorption coefficient aswell as computer generated values. Qualitatively,good agreement between the data compiled by Phil-ipp [50] and the other data is observed in Figs. 2–5.Moreover, residuals with the present model rangefrom �0.42 to 0.19 for n� and �0.25 to 0.37 for k� from7 to 50 �m.
5. Conclusions
A thorough review of experimental data for thecomplex refractive index of silica glass at near roomtemperatures over a spectral range from 30 nmto 1000 �m implies that the values reported in theliterature can vary significantly due to numeroussample features and experimental methods and con-ditions. Hence, it is essential to report the silica glasssynthesis method, composition, impurity, and defectslevel, sample thickness, surface roughness, and tem-peratures as well as the retrieval method and theunderlying assumptions when one reports opticalproperties of glass. However, the general features ofthe complex refractive index spectra are relativelyconsistent throughout the region of the spectrumconsidered. Silica glass is effectively opaque forwavelengths shorter than 200 nm and larger than3.5–4.0 �m. Strong absorption bands are observed (i)below 160 nm due to interaction with electrons, ab-sorption by impurities, and the presence of OHgroups and point defects, (ii) at �2.73–2.85, 3.5, and4.3 �m also caused by OH groups, (ii) at �9–9.5, 12.5,and 21–23 �m due to SiOOOSi resonance modes ofvibration. New formulas for both the real and imag-inary parts of the complex refractive index are pro-
8131 APPLIED OPTICS � Vol. 46, No. 33 � 20 November 2007
posed over a wide spectral range between 7 and50 �m, thus complementing the existing analyticalformula for n� in the range of 0.21–7 �m [16,23]. Theimaginary part of the complex refractive index canbe neglected in much of this range �0.21–4 �m�. Thedifferences between various experimental data arecomparable and greater than the differences betweenthe results of these formulas and the experimentaldata used to develop them. Hence it is believed thatthe formulas proposed are useful for practical engi-neering applications such as simulations and optimi-zations of optical and thermal systems. The datacollected and presented in this study are available indigital form online [79] or directly from the corre-sponding author upon request.
The authors thank Asahi Glass Co., Ltd. Japan forfinancial support. They are grateful to D. De SousaMeneses for helpful discussion about FOCUS. The con-tribution of M. Jonasz was supported by MJC OpticalTechnology.
References1. G. Hart, “The nomenclature of silica,” Am. Mineral. 12, 383–
395 (1927).2. R. B. Sosman, The Phase of Silica (Rutgers U. Press, 1964).3. G. Hetherington, K. H. Jack, and M. W. Ramsay, “The high-
temperature electrolysis of vitreous silica, part I. Oxidation,ultra-violet induced fluorescence, and irradiation colour,”Phys. Chem. Glasses 6, 6–15 (1965).
4. R. Bruckner, “Properties and structure of vitreous silica. I,” J.Non-Cryst. Solids 5, 123–175 (1970).
5. K. M. Davis, A. Agarwal, M. Tomozawa, and K. Hirao, “Quan-titative infrared spectroscopic measurement of hydroxyl con-centrations in silica glass,” J. Non-Cryst. Solids 203, 27–36(1996).
6. R. H. Doremus, Glass Science (Wiley, 1994).7. J. M. Senior, Optical Fiber Communications: Principles and
Practice, 2nd ed. (Prentice Hall, 1992).8. G. E. Keiser, Optical Fiber Communications, 3rd ed. (McGraw-
Hill Higher Education—International Editions: ElectricalEngineering Series, 2000).
9. B. Brixner, “Refractive-index interpolation for fused silica,” J.Opt. Soc. Am. 57, 674–676 (1967).
10. L. E. Sutton and O. N. Stavroudis, “Fitting refractive indexdata by least squares,” J. Opt. Soc. of Am. 51, 901–905 (1961).
11. T. Steyer, K. L. Day, and R. Huffman, “Infrared absorption bysmall amorphous quartz spheres,” Appl. Opt. 13, 1586–1590(1974).
12. T. Henning and H. Mutschke, “Low-temperature infraredproperties of cosmic dust analogues,” Astron. Astrophys. 327,743–754 (1997).
13. C. Koike, H. Hasegawa, N. Asada, and T. Komatuzaki,“Optical constants of fine particles for the infrared region,”Mon. Not. R. Astron. Soc. 239, 127–137 (1989).
14. M. Born and E. Wolf, Principles of Optics, 7th ed. (CambridgeU. Press, 1999).
15. C. M. Randall and R. D. Rawcliffe, “Refractive indices of ger-manium, silicon, and fused quartz in the far infrared,” Appl.Opt. 6, 1889–1895 (1967).
16. C. Tan, “Determination of refractive index of silica glass forinfrared wavelengths by IR spectroscopy,” J. Non-Cryst. Solids223, 158–163 (1998).
17. I. Fanderlik, “Glass science and technology,” Optical Propertiesof Glass (Elsevier Science, 1983), Vol. 5.
18. C. F. Bohren and D. R. Huffman, Absorption and Scattering ofLight by Small Particles (Wiley, 1983).
19. P. J. Riu and C. Lapaz, “Practical limits of the Kramers–Krönig relationships applied to experimental bioimpedancedata,” Ann. N.Y. Acad. Sci. 873, 374–380 (1999).
20. H. Philipp, “The infrared optical properties of SiO2 and SiO2
layers on silicon,” J. Appl. Phys. 50, 1053–1057 (1979).21. M. F. Modest, Radiative Heat Transfer (Academic, 2003).22. A. V. Dvurechensky, V. Petrov, and V. Y. Reznik, “Spectral
emissivity and absorption coefficient of silica glass at ex-tremely high temperatures in the semitransparent region,”Infrared Phys. 19, 465–469 (1979).
23. I. Malitson, “Interspecimen comparison of the refractive indexof fused silica,” J. Opt. Soc. Am. 55, 1205–1209 (1965).
24. A. M. Efimov, Optical Constants of Inorganic Glasses (CRC,1995).
25. L. Bogdan, “Measurement of radiative heat transfer with thin-film resistance thermometers,” NASA CR 27, 1–39 (1964).
26. A. Sviridova and N. Suikovskaya, “Transparent limits of in-terference films of hafnium and thorium oxides in the ultravi-olet region of the spectrum,” Opt. Spectrosc. 22, 509–512(1967).
27. E. Ellis, D. W. Johnson, A. Breeze, P. M. Magee, and P. G.Perkins, “The electronic structure and optical properties ofoxide glasses I. SiO2, Na2O:SiO2 and Na2O:CaO:SiO2,” Philos.Mag. B 40, 105–124 (1979).
28. C. Tan, “Optical interference and refractive index of silicaglass in the infrared absorption region,” J. Non-Cryst. Solids249, 51–54 (1999).
29. M. Khashan and A. Nassif, “Dispersion of the optical constantsof quartz and polymethyl methacrylate glasses in a wide spec-tral range: 0.2–3 �m,” Opt. Commun. 188, 129–139 (2001).
30. J. Reitzel, “Infrared spectra of SiO2 from 400 cm�1 to 600cm�1,” J. Chem. Phys. 23, 2407–2409 (1955).
31. G. M. Mansurov, R. K. Mamedov, S. Sudarushkin, V. K.Sidorin, K. K. Sidorin, V. I. Pshenitsyn, and V. M. Zolotarev,“Study of the nature of a polished quartz-glass surface byellipsometric and spectroscopic methods,” Opt. Spectrosc. 52,852–857 (1982).
32. Y. Ikuta, S. Kikugawa, T. Kawahara, H. Mishiro, N. Shimo-daira, and S. Yoshizawa, “New silica glass AQF for 157-nmlithography,” Proc. SPIE 4000, 1510–1514 (2000).
33. K. Kajihara, “Improvement of vacuum-ultraviolet transpar-ency of silica glass by modification of point defects,” J. Ceram.Soc. Jpn. 115, 85–91 (2007).
34. G. H. Sigel, “Ultraviolet spectra of silicate glasses: a review ofsome experimental evidence,” J. Non-Cryst. Solids 13, 372–398 (1974).
35. N. Shimodaira, K. Saito, A. Ikushima, T. Kamihori, and S.Yoshizawa, “UV transmittance of fused silica glass influencedby thermal disorder,” Proc. SPIE 4000, 1553–1559 (2000).
36. H. Rawson, “Glass science and technology,” Properties andApplications of Glass (Elsevier Science, 1980), Vol. 3.
37. C. M. Smith and L. A. Moore, “Fused silica for 157-nm trans-mittance,” Proc. SPIE 3676, 834–841 (1999).
38. Y. Ikuta, S. Kikugawa, T. Kawahara, H. Mishiro, K. Okada, K.Ochiai, K. Hino, T. Nakajima, M. Kawata, and S. Yoshizawa,“New modified silica glass for 157-nm lithography,” Proc. SPIE4066, 564–570 (2000).
39. D. Griscom, “The electronic structure of SiO2: a review of re-cent spectroscopic and theoretical advances,” J. Non-Cryst.Solids 24, 155–234 (1977).
40. V. Petrov and S. Stepanov, “Radiation characteristics of quartzglasses spectral radiating power,” Teplofiz. Vys. Temp. 13,335–345 (1975).
41. V. Plotnichenko, V. Sokolov, and E. Dianov, “Hydroxyl groupsin high-purity silica glass,” J. Non-Cryst. Solids 261, 186–194(2000).
20 November 2007 � Vol. 46, No. 33 � APPLIED OPTICS 8132
42. A. M. Efimov and V. G. Pogareva, “IR absorption spectra ofvitreous silica and silicate glasses: The nature of bands in the1300 to 5000 cm�1 region,” Chem. Geol. 229, 198–217 (2006).
43. D. J. Griffiths, Introduction to Electrodynamics, 3rd ed. (Pren-tice Hall, 1999).
44. W. Rodney and R. Spindler, “Index of refraction of fused quartzfor ultraviolet, visible, and infrared wavelengths,” J. Opt. Soc.Am. 44, 677–679 (1954).
45. C. Tan and J. Arndt, “Temperature dependence of refractiveindex of glass SiO2 in the infrared wavelength range,” J. Phys.Chem. Solids 61, 1315–1320 (2000).
46. L. Dombrovsky, “Quartz-fiber thermal insulation: infrared ra-diative properties and calculation of radiative-conductive heattransfer,” J. Heat Transfer 118, 408–414 (1996).
47. D. D. S. Meneses, G. Gruener, M. Malki, and P. Echegut,“Causal Voigt profile for modeling reflectivity spectra ofglasses,” J. Non-Cryst. Solids 351, 124–129 (2005).
48. D. D. S. Meneses, M. Malki, and P. Echegut, “Structure andlattice dynamics of binary lead silicate glasses investigated byinfrared spectroscopy,” J. Non-Crystal. Solids 352, 769–776(2006).
49. S. Popova, T. Tolstykh, and V. Vorobev, “Optical characteris-tics of amorphous quartz in the 1400–200 cm�1 region,” Opt.Spectrosc. 33, 444–445 (1972).
50. H. R. Philipp, “Silicon dioxide (SiO2) glass,” in Handbook ofOptical Constants of Solids, E. D. Palik, ed. (Academic, 1985),Vol. I, pp. 749.
51. D. G. Drummond, “The infra-red absorption spectra of quartzand fused silica from 1 to 7.5 � II—experimental results,” Proc.R. Soc. London Ser. A 153, 328–339 (1936).
52. H. R. Philipp, “Optical transitions in crystalline and fusedquartz,” Solid State Commun. 4, 73–75 (1966).
53. H. R. Philipp, “Optical properties of non-crystalline Si, SiO,SiOx and SiO2,” J. Phys. Chem. Solids 32, 1935–1945 (1971).
54. P. Lamy, “Optical constants of crystalline and fused quartz inthe far ultraviolet,” Appl. Opt. 16, 2212–2214 (1977).
55. C. Tan and J. Arndt, “Refractive index, optical dispersion, andgroup velocity of infrared wave in silica glass,” J. Phys. Chem.Solids 62, 1087–1092 (2001).
56. C. Boeckner, “A method of obtaining the optical constants ofmetallically reflecting substances in the infrared,” J. Opt. Soc.Am. 19, 7–15 (1929).
57. O. Girin, Y. Kondratev, and E. Raaben, “Optical constants andspectral microcharacteristics of NaO2–SiO2 glasses in the IRregion of the spectrum,” Opt. Spectrosc. 29, 397–403 (1970).
58. J. Wray and J. Neu, “Refractive index of several glasses as afunction of wavelength and temperature,” J. Opt. Soc. Am. 59,774–776 (1969).
59. V. Zolotarev, “The optical constants of amorphous SiO2 andGeO2 in the valence band region,” Opt. Spectrosc. 29, 34–37(1970).
60. I. Simon and H. McMahon, “Study of the structure of quartz,cristobalite, and vitreous silica by reflection in infrared,”J. Chem. Phys. 21, 23–30 (1953).
61. M. Herzberger and C. Salzberg, “Refractive indices of infraredoptical materials and color correction of infrared lenses,” J.Opt. Am. 52, 420–427 (1962).
62. T. Yamamuro, S. Sato, T. Zenno, N. Takeyama, H. Matsuhara,I. Maeda, and Y. Matsueda, “Measurement of refractive indi-ces of 20 optical materials at low temperatures,” Opt. Eng. 45,083401 (2006).
63. H. Bach and N. Neuroth, eds., The Properties of Optical Glass,2nd ed. (Springer-Verlag, 2004).
64. E. Beder, C. Bass, and W. Shackleford, “Transmittivity andabsorption of fused quartz between 0.2 and 3.5 �m from roomtemperature to 1500 °C,” J. Am. Ceram. Soc. 10, 2263–2268(1971).
65. D. Gillespie, A. Olsen, and L. Nichols, “Transmittance of opti-cal materials at high temperatures in the 1-� to 12-� range,”Appl. Opt. 4, 1488–1493 (1965).
66. G. Calingaert, S. Heron, and R. Stair, “Sapphire and other newcombustion-chamber window materials,” SAE J. 39, 448–450(1936).
67. D. Heath and P. Sacher, “Effects of a simulated high-energyspace environment on the ultraviolet transmittance of opticalmaterial between 1050 Å and 3000 Å,” Appl. Opt. 5, 937–943(1966).
68. A. F. Grenis and M. J. Matkovich, “Blackbody reference fortemperature above 1200 K. Study for design requirements,”AMRA TR 65, 1–18 (1965).
69. G. V. Saidov and E. B. Bernstein, “Optical constants of surfacelayer of fused quartz in the 900–1300 cm�1 range,” Fiz. Khim.Stekla 8, 75–81 (1982).
70. A. M. Efimov, “Dispersion of optical constants of vitreoussolids,” Ph.D. dissertation (Vavilov State Optical Institute,Leningrad).
71. R. K. Bogens and A. G. Zhukov, “The optical constants of fusedquartz in the far infrared,” J. Appl. Spectrosc. 25, 54–55(1966).
72. T. J. Parker, J. E. Ford, and W. G. Chambers, “The opticalconstants of pure fused quartz in the far-infrared,” InfraredPhys. 18, 215–219 (1978).
73. A. P. Zhilinskii, A. P. Gorchakov, T. S. Egorova, and N. A.Miskinova, “Optical characteristics of fused quartz in the farIR range,” Opt. Spectrosc. 62, 783–784 (1987).
74. W. Bagdade and R. Stolen, “Far infrared absorption in fusedquartz and soft glass,” J. Phys. Chem. Solids 29, 2001–2008(1968).
75. G.-L. Tan, M. F. Lemon, and R. H. French, “Optical propertiesand London dispersion forces of amorphous silica determinedby vacuum ultraviolet spectroscopy and spectroscopic ellip-sometry,” J. Am. Ceram. Soc. 86, 1885–1892 (2003).
76. P. T. T. Wong and E. Whalley, “Infrared and Raman spectra ofglasses. Part 2. Far infrared spectrum of vitreous silica in therange 100–15 cm�1,” Discuss. Faraday Soc. 50, 94–102 (1970).
77. R. K. Mamedov, G. M. Mansurov, and N. I. Dubovikov, “Opti-cal constants of quartz glass in the IR range,” Opt. Mekh.Prom. 4, 56 (1982) [Sov. J. Opt. Technol. 49, 256 (1982)].
78. M. Miler, “Infrared absorption of glassy silicon dioxide,” Czech.J. Phys. 18, 354–362 (1968).
79. http://www.tpdsci.com/Tpc/RIQtzFsd.php and http://repositories.cdlib.org/escholarship/.
8133 APPLIED OPTICS � Vol. 46, No. 33 � 20 November 2007