Refractive index dispersion of doped silica for fiber optics

Fiber Optics Research Center, Russian Academy of Sciences, Moskva, Moscow, Russia
Optics Communications (Impact Factor: 1.45). 12/2002; 213(4-6):301-308. DOI: 10.1016/S0030-4018(02)02087-4
The spectral dependencies of refractive index have been measured in Ge-, P-, N-, Cl-, B-, F-, and Al-doped silica glasses as well as in undoped silica glasses using bulk prism samples cut from optical fiber preforms. The latter have been fabricated by MCVD-, PCVD-, and SPCVD-processes. Based on the experimental results, material dispersion in the glasses has been analyzed, which is one of the most important parameters for fiber optics. An assumption has been made regarding the origin of the significant discrepancy of the zero-dispersion wavelength of nominally identical glass compositions in different publications. The effect of chlorine admixture on the dispersion curves has been investigated. Nitrogen-doped silica is shown to be a promising material for broadband graded-index multimode fibers.
Refractive index dispersion of doped silica for fiber optics
O.V. Butov
, K.M. Golant
, A.L. Tomashuk
, M.J.N. van Stralen
A.H.E. Breuls
Fiber Optics Research Center, Plasma Chemical Laboratory, General Physics Institute, Russian Academy of Sciences,
38 Vavilov Street, Moscow 119991, Russia
Draka Comteq, Draka Fibre Technology B.V., Eindhoven, The Netherlands
Received 10 July 2002; received in revised form 19 September 2002; accepted 14 October 2002
The spectral dependencies of refractive index have been measured in Ge-, P-, N-, Cl-, B-, F-, and Al-doped silica
glasses as well as in undoped silica glasses using bulk prism samples cut from optical fiber preforms. The latter have
been fabricated by MCVD-, PCVD-, and SPCVD-processes. Based on the experimental results, material dispersion in
the glasses has been analyzed, which is one of the most important parameters for fiber optics. An assumption has been
made regarding the origin of the significant discrepancy of the zero-dispersion wavelength of nominally identical glass
compositions in different publications. The effect of chlorine admixture on the dispersion curves has been investigated.
Nitrogen-doped silica is shown to be a promising material for broadband graded-index multimode fibers.
Ó 2002 Elsevier Science B.V. All rights reserved.
PACS: 78.20.Ci; 42.81.Ht
Keywords: Refractive index dispersion; Chromatic dispersion; Material dispersion; Inter-modal dispersion; Multimode fiber; Nitrogen-
doped silica fiber
1. Introduction
One of the basic parameters, characterizing a
given glass as an optical material is the refractive
index dispersion. This parameter is of a particular
importance in complex optical systems, in which
aberration due to dispersion is to be fully mini-
mized (e.g., in optical telescopes), or where the
precise value of dispersion is to be known in a
wide spectral range (e.g., in prism spectrum
Refractive index dispersion is of major impor-
tance in fiber optic telecommunications. In par-
ticular, the information capacity of a single-mode
fiber depends primarily on its chromatic disper-
sion. The latter consists of waveguide and material
dispersion. The waveguide dispersion can be ad-
justed by choosing an appropriate refractive index
profile of the fiber, whereas, the material disper-
Optics Communications 213 (2002) 301–308
Corresponding author. Tel.: +7-095-132-8303; fax: +7-095-
E-mail address: (O.V. Butov).
0030-4018/02/$ - see front matter Ó 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 030-401 8 ( 0 2 ) 0 2 0 8 7 - 4
Page 1
sion is an invariable characteristic of a specific
material (e.g., see [1]). Therefore, reliable and
accurate data on the material dispersion is in-
dispensable in designing high-bit-rate optical
communication systems.
M ¼
: ð1Þ
Multimode fibers are usually used in fiber-optic
local area networks. In this case, inter-modal dis-
persion is the main factor limiting the information
capacity (e.g., see [2]). A graded-index profile of
the fiber core is used to minimize this dispersion.
However, the optimal profile shape also depends
on refractive index dispersion of the core and
cladding materials and is defined by formula (2)
nðr; kÞ¼n
ðkÞ 1 2D
r < a
nðr; kÞ¼n
r P a;
where n
is the maximum value of core refractive
index, n
the refractive index of cladding,
D ¼ n
, r the distance from the fiber
center, and a is the core radius;
¼ 2 þ y D
ð4 þ yÞð3 þ yÞ
ð5 þ 2yÞ
; ð3Þ
y ¼
; ð4Þ
and N
is the material group index:
¼ n
: ð5Þ
The last term in (3) is small and can be neglected.
As follows from (2)–(5), optimal profile pa-
rameter a
depends on the dispersion character-
istics of core and cladding materials and varies
with wavelength. Hence, it remains to be an acute
task to find an optimal core material for which the
parameter would not significantly depend on
wavelength in a wide spectral region. Finding a
single optimal profile for a wide spectral region
would make it possible to use one and the same
fiber at different wavelengths without a reduction
of the bit-rate due to inter-modal dispersion. It
would permit application of a practically unlimited
number of spectral channels and, hence, a dra-
matic increase of the information capacity of
optical communication links. Thus, accurate de-
termination of the material dispersion of the fiber
glass is a topical problem.
Germanium is currently the most important
dopant when forming the core of the silica fiber [4].
Apart from germanium some other dopants are
used, e.g., phosphorus, fluorine, boron, and alu-
minum [3,5,6]. A comparatively new dopant is
nitrogen [7]. In many ways nitrogen-doped-silica-
core fibers are quite competitive with german-
osilicate fibers. The former are superior to
germanosilicate fibers in a number of parameters,
such as radiation resistance and thermal resistance
of in-fiber Bragg gratings [8–10]. Among the ad-
vantages of nitrogen-doped silica-core fibers is
their potential low cost, because nitrogen is far
cheaper than germanium and its resources cannot
be exhausted.
Dispersion properties of doped glasses for fiber
optics have been studied during the last 30 years;
however, the results of different authors vary
greatly (e.g., see [2,11]). This research work is
aimed at clarifying the disputable points related to
dispersion of glasses for fiber optics by means of
increasing the accuracy and spectral range of the
dispersion measurements. In addition to tradi-
tional dopants to silica, we investigate nitrogen
and chlorine, the latter being an intrinsic techno-
logical admixture present in all optical fibers.
We also compare data on undoped silica, synthe-
sized by the modified chemical vapor deposi-
tion (MCVD) [12], plasma-activated chemical
vapor deposition (PCVD) [13], surface plasma
chemical vapor deposition (SPCVD) [14] technol-
2. Experiment
The nðkÞ dependence was measured by the
classic method of white light decomposition on a
prism cut from the material under study. The op-
tical scheme of the experimental set-up is shown in
Fig. 1. A parallel white light beam from a halogen
lamp was collimated by elliptic and parabolic
mirrors and then fell on the prism. The decom-
302 O.V. Butov et al. / Optics Communications 213 (2002) 301–308
Page 2
posed light passed through a monochromator and
was registered by a photoreceiver installed behind
the exit slit of the monochromator. The prism was
placed on a rotational micropositioner and the
light wavelength incident on the monochromator
entrance slit varying in rotating the prism (Fig. 2).
The measuring procedure consisted in finding the
prism rotation angle ensuring the maximal signal
at the wavelength set by the monochromator.
Knowing this angle, one can calculate the refrac-
tive index of the material at a given wavelength
(see formula (6))
¼ arcsinðsinði
h ¼ i
þ i
¼ i
a þ r
: ð6Þ
All the angles in (6) are shown in Fig. 2.
The measurements were carried out in the
wavelengths range 0.3–2.6 lm with a step of 10–50
nm. As photoreceivers we used a photomultiplier
for wavelengths 0.3–0.8 lm and a cooled InSb
photodiode for wavelengths 0.8–2.6 lm. The
photoreceiver signal was processed by a lock-in.
The lock-in, the rotational micropositioner, and
the monochromator were controlled by a com-
puter. Thus, the measuring procedure was fully
The experimental set-up allowed measuring
nðkÞ to an accuracy better than 10
. For further
data processing, in particular for defining material
dispersion and a
from relations (1) and (3), the
experimental points were approximated by three-
term Selmeier formula (7) [4,15].
ðkÞ¼1 þ
; ð7Þ
where k is the wavelength, n the refractive index as
a function of k, a
the oscillator strength, and b
the oscillator resonance frequency.
Calculation of the derivatives based on approx-
imation (7) depends on the spectral interval within
which the measurements have been performed. For
example, let us consider what would happen if we
would not take into account some experimental
points for sample 1 available for the spectral range
0.4–2:0 lm. Removing the experimental points in
the ranges between 0.3–0.4 and 2.05–2:6 lm and
thus narrowing the spectral interval lead us to the
following values for Sellmeier coefficients:
¼ 0:9415; b
¼ 0:07478; a
¼ 0:16251; b
0:14261; a
¼ 0:51434; b
¼ 7:64773. These values
significantly differ from those listed in the table.
Obviously the derivatives of the nðkÞ function will
differ too. That is why it is very important to widen
the spectral range in order to improve the accuracy
of the derivatives of the nðkÞ dependence.
Fig. 2. Light dispersion in a prism.
Fig. 1. Optical diagram of the experimental set-up.
O.V. Butov et al. / Optics Communications 213 (2002) 301–308 303
Page 3
Glass specimens studied are listed in Table 1.
The prisms with a vertex angle of 30° were made of
fiber preforms. The measurements were conducted
in the working zone composed of the doped core
of the preform (Fig. 3). The working zone was set
aside by a diaphragm on the lateral prism surface.
The core radius varied from sample to sample in
Table 1
Experimental samples characterization
of sample
Doping elements and
its concentration
Dn Sellmeier
coefficients a
coefficients b
1 MCVD Cl (0.06 wt%) 0.50716 0.04014
0.59707 0.11359
0.69879 8.81674
2 SPCVD Cl (0.05 wt%) 0.70209 0.05714
0.40292 0.1239
1.60979 12.95448
3 PCVD Cl (0.3 wt%) 0.88671 0.07954
0.21675 0.1244
0.69401 8.83315
4 SPCVD Cl (3.4 wt%) 0.0029 0.53502 0.04792
0.577 0.11548
0.65548 8.63483
5 PCVD F (0.9 wt%) )0.0068 0.87219 0.07417
Cl (0.13 wt%) 0.21238 0.1298
0.94959 10.22611
6 SPCVD B (out of range of the spectrometer) )0.0014 0.54956 0.03566
Cl (0.06 wt%) 0.55151 0.11823
1.52791 12.54703
7 SPCVD Al (4.9 wt%) 0.009 0.91249 0.08088
Cl (0.06 wt%) 0.21688 0.12558
0.77945 9.39992
(4.5 mol.%) 0.0062 0.49211 0.04807
Cl (0.16 wt%) 0.62925 0.11275
0.59202 8.29299
(11.6 mol.%) 0.0166 0.49795 0.04407
Cl (0.16 wt%) 0.65295 0.11754
0.83515 9.86362
10 MCVD P (12.5 wt%) 0.0135 0.51512 0.02636
Cl (0.03 wt%) 0.62804 0.11614
1.0743 10.6931
11 SPCVD N (1.6 wt%) 0.0151 0.49798 0.05043
Cl (0.9 wt%) 0.64994 0.11155
1.39632 12.14576
Fig. 3. Lateral view of a prism used in the experiments.
304 O.V. Butov et al. / Optics Communications 213 (2002) 301–308
Page 4
the range of 1–5 mm, inhomogeneity of the sam-
ples being less than 10
3. Results and discussion
The result of the nðkÞ measurement in undoped
silica fabricated by the MCVD method is pre-
sented in Fig. 4. The solid line shows the approx-
imation of the experimental points with the help of
formula (7).
As is seen from Fig. 4, three-term Sellmeier re-
lation (7) well approximates the experimental data
in the spectral region under consideration. Unlike
[16], where nðkÞ calculations were made from indi-
rect measurements, no abnormality in the disper-
sion curve behavior, at least within the experimental
error of less than 10
, was observed. Therefore, by
contrast to [16] our data confirms applicability of
formula (7) for approximation of nðkÞ curves in the
region of high transparency.
The wavelength range in the vicinity of zero dis-
persion k
Þ¼0Þ, where temporal broadening
of a signal propagating through a fiber is nearly
minimal, is of much interest for fiber optics [11]. As
was discussed above, dispersion in a single-mode
fiber is primarily determined by the dispersion
properties of the core material. Let us consider the
results of the material dispersion measurements in
various glasses for fiber optics. Fig. 5 shows M ðkÞ
plots for various silicas calculated from the experi-
mental data by formula (1).
First, let us compare undoped silicas (Fig. 5(a))
fabricated by different technologies. As follows
from Fig. 5(a), the values of k
for the MCVD and
SPCVD glasses practically coincide; however, for
the PCVD silica, k
is shifted to longer wave-
lengths. This can be due to different chlorine
content in the glasses. According to the chemical
analysis, the former two glasses contained 005–
0.06 wt% of chlorine, while the latter contained 0.3
wt%. To check the assumption of the role of
chlorine, we made a glass containing 3.4 wt% of
chlorine using the SPCVD technology. It was
found that the presence of chlorine in the glass did
shift the zero dispersion wavelength to longer
wavelengths (Fig. 5(a)). However, this shift turned
out to be disproportionately great for the PCVD
undoped silica. Note that literature values of the
zero dispersion wavelength of undoped silica also
Fig. 4. Refractive index dispersion of undoped MCVD silica
Fig. 5. Material dispersion of the glass samples.
O.V. Butov et al. / Optics Communications 213 (2002) 301–308 305
Page 5
vary in the range 1.26–1:285 lm [6,11,17]. In all
likelihood, it is peculiarities of the technological
process (including deposition of silica on the inner
surface of a substrate tube and its collapsing (e.g.,
see [12]) that influence the position of the zero
dispersion wavelength. For example, technology-
dependent point defects, such as oxygen-deficient
centers or fabrication-induced stresses, can signif-
icantly influence the behavior of the dispersion
curve. Although stress in fibers and fiber preforms
may be rather different, we will not discuss here the
influence of this important effect on fiber disper-
Now let us consider the behavior of zero
dispersion wavelength in doped glasses for fiber
optics (Table 1, Fig. 5). The k
value of the fluo-
rine-doped glass practically coincides with that of
undoped silica synthesized by the MCVD tech-
nology. Our result agrees with the data [6]; how-
ever, as well as in the case of undoped silica, the
results obtained in different papers significantly
vary [6,17].
Boron and fluorine incorporation in the glass
decreases refractive index. But unlike fluorine,
boron strongly shifts k
to shorter wavelengths.
This property, as is known from [2], can be used to
compensate the dispersion alteration due to ger-
manium doping.
Aluminum is usually applied in erbium-doped
active fiber. Al considerably influences the disper-
sion properties of silica, shifting k
to longer
wavelengths. These data are in agreement with the
results of previous research [3].
Germanium-doped silica is currently the most
important material for fiber optics. We examined
two specimens of germanosilicate glass obtained
by the PCVD technology with different Ge content
(Fig. 5(b)). In Fig. 6, the k
dependence on the
amount of germanium dioxide in the glass is de-
picted. For comparison, the results of other papers
[2,11,17,18] are given in the same plot. The solid
line is a linear approximation of our results and
those taken from the literature. As follows from
Fig. 6, the results on germanosilicate glass con-
siderably vary. This variation can be attributed to
the peculiarities of the specific fabrication tech-
nologies, which might result in different concen-
tration of chlorine and point defects, such as
germanium oxygen-deficient centers (GODCs).
The above factors are usually not taken into ac-
count in interpreting the refractive index disper-
sion in silica glasses. In particular, it is known that
the GODCs concentration in glasses with the same
germanium content can vary by four orders of
magnitude [19] depending on the specific technol-
ogy. It is likely that the concentration of point
defects is the major cause of the k
discrepancy in
different papers, because variations in the chlorine
concentration cannot result in such a considerable
shift (see Fig. 5(a)). The influence of specific
point defects on the glass dispersion properties
calls for further study.
Phosphorus-doped silica is also widely used in
fiber optics and optoelectronics. Phosphorus in-
corporation in silica increases its refractive index
practically without alteration of dispersion. It
follows from [17] that k
in phosphorus-doped
silica is insignificantly shifted to shorter wave-
lengths. According to our results, k
shifts to
longer wavelengths, as compared to undoped
MCVD silica, and to shorter wavelengths as
compared to undoped silica fabricated by PCVD
(Fig. 5(a)). Generally, it can be concluded that
material dispersion of phosphorus-doped silica is
similar to that of undoped silica. Hence, a unique
property of phosphorus-doped silica, as follows
from formula (2), is that it can be used for fabri-
cation of wide-band multimode fibers with mini-
mal inter-modal dispersion in a broad spectral
Fig. 6. Comparative data on zero material dispersion wave-
length of Ge-doped glasses. The solid line is a linear fit of our
results and those taken from the literature.
306 O.V. Butov et al. / Optics Communications 213 (2002) 301–308
Page 6
range. Fig. 7 shows the dependence of the optimal
refractive index profile parameter a
, calculated
by formula (2), on wavelength for variously doped
glasses for fiber optics. A glass composition can be
considered as optimal, if a
is practically un-
changed in a wide spectral range. It is seen (Fig. 7)
that the calculated index profile parameter a
varies only slightly in the case of phosphorus-
doped silica, from 2.02 to 2.05, in the spectral
range 0.7–1:7 lm. This fact makes it possible to
fully cover the currently existing spectral telecom
windows using a unified refractive index shape of
multimode graded-index fibers regardless of the
specific spectral window.
Nitrogen-doped silica is of particular interest
(Fig. 5(a) and (b)). Even a small addition of nitrogen
strongly increases the refractive index of silica
(Table 1). Therewith, nitrogen doping does not
significantly alter the material dispersion of silica
(Fig. 5(b)). It should be noted that chlorine admix-
ture also influences the dispersion of nitrogen-
doped silica (Table 1). Thus, if we take into account
the chlorine contribution, the dispersion alteration
caused immediately by nitrogen incorporation will
turn out to be minimal. In other words, as a result of
nitrogen doping the refractive index of silica in-
creases evenly in a wide spectral region. This fact
opens up possibilities for using nitrogen-doped sil-
ica as the core material of wide-band graded-index
fibers. As follows from formula (2), the fiber clad-
ding material considerably influences the a
Until recently, multimode graded-index fibers with
an undoped silica cladding and phosphorus-doped
silica core were known to be the only candidates for
applications in wide spectral ranges. As is seen from
Fig. 7, a
for nitrogen-doped-silica-core undoped-
silica-cladding fibers does not strongly depend on
wavelength either. For comparison, similar depen-
dencies are also depicted for such dopants as ger-
manium and chlorine. When finding optimal
technology, it is necessary to keep in mind a con-
siderable influence of chlorine contained in both
core and cladding of the fiber, as well as an influence
of possible point defects, on a
. To illustrate such
possible effects, Fig. 7 shows a
ðkÞ dependences for
identical nitrogen-doped-silica-core fibers with an
undoped silica cladding fabricated by different
methods. Thus, by optimizing the chlorine concen-
tration in the core of graded-index nitrogen-doped
silica fibers with regard to the dispersion properties
of the undoped cladding, one can minimize the in-
ter-modal dispersion in the fibers simultaneously in
the whole range from 0.7 to 1:7 lm.
4. Conclusion
The refractive index as a function of wavelength
was directly measured to a high accuracy in a wide
spectral range for standard and new silicas for fi-
ber optics. The results of the measurements were
used to analyze the material dispersion of these
glasses. The influence of the technological chlorine
impurity on the dispersion curves was revealed.
The dispersion parameters of new materials for
fiber optics, e.g., nitrogen-doped silica, were de-
termined. In particular, it was found that nitrogen-
doped silica is a well-suited material for wide-band
graded-index fibers. It was shown that the disper-
sion properties of such fibers can be easily opti-
mized for any type of silica used in the cladding.
The authors are grateful to Dr. M.M. Bubnov
for MCVD specimens supply and to Dr. V.M.
Mashinsky for a fruitful discussion of the results
Fig. 7. Optimal a-profile parameter as a function of wavelength
for multimode graded-index fibers of different glass composi-
O.V. Butov et al. / Optics Communications 213 (2002) 301–308 307
Page 7
[1] S.H. Wemple, Appl. Opt. 18 (1979) 31.
[2] J.W. Fleming, J. Am. Ceram. Soc. 59 (1976) 503.
[3] H.M. Presby, I.P. Kaminow, Appl. Opt. 15 (1976) 3029.
[4] J.W. Fleming, Appl. Opt. 23 (1984) 4486.
[5] C.R. Hammond, S.R. Norman, Opt. Quant. Electron.
9 (1977) 399.
[6] J.W. Fleming, D.L. Wood, Appl. Opt. 22 (1983) 3102.
[7] E.M. Dianov, K.M. Golant, R.R. Khrapko, A.S. Kur-
kov, A.L. Tomashuk, J. Lightwave Technol. 13 (1995)
[8] E.M. Dianov, K.M. Golant, R.R. Khrapko, A.L. Toma-
shuk, Electron. Lett. 31 (1995) 1490.
[9] E.M. Dianov, K.M. Golant, R.R. Khrapko, A.S. Kurkov,
B. Leconte, M. Douay, P. Bernage, P. Niay, Electron. Lett.
33 (1997) 236.
[10] O.V. Butov, K.M. Golant, I.V. Nikolin, Electron. Lett.
38 (2002) 523.
[11] V.M. Mashinskiy, Proc. Inst. Gen. Phys. Acad. Sci. USSR
5 (1988) 111.
[12] J. Kirchhof, Cryst. Res. Technol. 20 (1985) 705.
[13] G. Kuyt, B.A.M. Teunissen, Philips TDS Rev. 46 (1988) 1.
[14] D. Pavy, M. Moisan, S. Saada, P. Chollet, P. Leprince, J.
Marrec, in: Proc. 12th European Conf. Opt. Commun.,
Barcelona, 1986, p. 19.
[15] W. Hermann, D.U. Wiechert, Mater. Res. Bull. 24 (1989)
[16] M.A. Khashan, A.Y. Nassif, Opt. Commun. 188 (2001) 129.
[17] J.W. Fleming, Electron. Lett. 14 (1978) 326.
[18] D.L. Wood, J.W. Fleming, Rev. Sci. Instrum. 53 (1982) 43.
[19] K.M. Golant, O.V. Butov, A.N. Denisov, V.M. Mashinsky,
O.D. Sazhin, C.M. Smith, S.V. Muraviov, Phys. Chem.
Glasses 43c (2002) 131.
308 O.V. Butov et al. / Optics Communications 213 (2002) 301–308
Page 8
  • Source
    • "From an inset in Fig. 6, one can also see that, for long-lasting (>5 sec) UV exposure durations, a significant growth of SMC-related absorption around 330 nm with an increase in the chlorine concentration is observed. One can assume that an increase in the chlorine concentration should lead to an increase in the amount of inhomogeneities in the glass bulk that occur in the form of discontinuities of the glass network [24]. In this case, conditions for the formation of neutral silver molecular clusters should become more favorable and, consequently, their concentration should increase. "
    Full-text · Article · May 2016 · Optical Materials Express
  • Source
    • "However, 10% La dopant decreases the refractive index (Figure 6). Previous researches [68, 69] showed that the refractive index tends to increase with the increasing in dopant concentration, which is caused by the occurrence of disorder in structure, changes in stoichiometry, and internal strain caused by polarizability. Consequently, if containing dopant materials have high polarizability, the refractive index will thus increase [70]. "
    [Show abstract] [Hide abstract] ABSTRACT: Undoped, 5%, and 10% lanthanum doped lithium tantalate thin films were annealed at 550°C temperature for 12.5 hours and their properties were characterized. The results showed that the dopant addition affects the crystal formation of LiTaO3, especially the 10% La2O5 dopant, where the lattice of a = 5.11 Å, c = 13.30 Å, and its crystal was hexagonal. The observed functional groups were O-H, N-H, C = C, Li-O, Ta-O. The higher La2O5 dopant concentration leads to the higher absorption of Ta-O and Li-O, the lower the energy gap, the higher the refractive index, and the smaller the particle size of thin films.
    Full-text · Article · Nov 2015 · Integrated Ferroelectrics
  • Source
    • "Exposure to 244 nm light ∆n mod ≈ 10 −3 in unloaded fiber ∆n mod ≈ 10 −2 in H 2 loaded fiber Exposure of the fiber to 244 nm pulsed light leads to a poor photosensitivity in either a H 2 -loaded or unloaded fiber (≈3 × 10 −5 at a cumulated fluence of 15 kJ/cm 2 ). Type II A photosensitivity in unloaded fiber exposed to 193 nm pulsed light (≈10 −3 ) [120,138,173] Doping a silica glass by PbO changes in refractive index are expected to be due to local modifications to the glass structure. ∆n mod = 0.21 at 633 nm λ p = 266 nm, 25 mJ/cm 2 per pulse, cumulated fluence of 150 J/cm 2 SF59 Schott glass surface relief and refractive index gratings (α UV ≈ 0.01 mm −1 in the glass at 266 nm). "
    [Show abstract] [Hide abstract] ABSTRACT: Over the last two decades UV-induced ΔnΔn profiling in SiO2 glasses was widely used for production of in-fibre/waveguide Bragg grating-based (BG) optical devices for photonics industry. These devices have found numerous applications in optical fiber sensing, telecommunication and all fiber laser systems. From a practical point of view, it is the most important photo-induced phenomenon observed each time a silica glass is exposed to convenient low or high UV intensity laser light through one quantum and multi-photon mechanisms respectively. In fact, depending on the materials, conditions of exposure and conditioning processes (i.e. the photosensitization process), UV induced index changes may vary from 10−5 up to 10−2. In the following, for the purpose of illustrating the complexity of this multiple-variable dependence, we present a review of how factors such as exposure time, laser wavelength, sensitization process, pulse duration, or the chemical composition, can affect the photosensitive response of silica-based glasses, i.e. the number of photons involved in the initial step of absorption, the writing efficiency and so on.
    Full-text · Article · Feb 2013 · Physics Reports
Show more