Depletion region in thermally poled fused silica
A. L. C. Triques, C. M. B. Cordeiro, V. Balestrieri, B. Lesche, W. Margulis, and I. C. S. Carvalho
Citation: Applied Physics Letters 76, 2496 (2000); doi: 10.1063/1.126387
View online: http://dx.doi.org/10.1063/1.126387
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/76/18?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Thermal poling induced second-order nonlinearity in femtosecond-laser-modified fused silica
Appl. Phys. Lett. 93, 061115 (2008); 10.1063/1.2973149
Thermal poling and ultraviolet erasure characteristics of type-III ultraviolet-grade fused silica and application to
periodic poling on planar substrates
J. Appl. Phys. 94, 1531 (2003); 10.1063/1.1591076
Time evolution of depletion region in poled silica
Appl. Phys. Lett. 82, 2948 (2003); 10.1063/1.1572472
Dynamics of the second-order nonlinearity in thermally poled silica glass
Appl. Phys. Lett. 79, 2687 (2001); 10.1063/1.1394948
Secondary ion mass spectrometry study of space-charge formation in thermally poled fused silica
J. Appl. Phys. 86, 6634 (1999); 10.1063/1.371736
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
143.106.108.174 On: Fri, 26 Jun 2015 14:12:57
APPLIED PHYSICS LETTERS
VOLUME 76, NUMBER 18
1 MAY 2000
Depletion region in thermally poled fused silica
A. L. C. Triques,a) C. M. B. Cordeiro,b) and V. Balestrieri
Departamento de Fı́sica, Pontifı́cia Universidade Católica do Rio de Janeiro,
22452-970 Rio de Janeiro, Brazil
B. Lesche
Departamento de Fı́sica, Universidade Federal de Juiz de Fora, 36036-330 Juiz de Fora, Brazil
W. Margulisc) and I. C. S. Carvalho
Departamento de Fı́sica, Pontifı́cia Universidade Católica do Rio de Janeiro,
22452-970 Rio de Janeiro, Brazil
共Received 1 February 2000; accepted for publication 9 March 2000兲
The depletion-layer width and the recorded electric field in thermally poled fused silica are
investigated experimentally as a function of the applied voltage. The depletion-layer width is
observed to vary linearly with the poling voltage. The average electric field recorded in the depletion
region was found to be (5.3⫾0.3)⫻108 V/m for all samples, independently of the poling voltage.
© 2000 American Institute of Physics. 关S0003-6951共00兲03218-6兴
Silica-based glass is one of the most important materials
in telecommunications. Nonlinear features in silicate glasses
are desirable in active devices. The macroscopic inversion
symmetry prevents even-order nonlinear effects from taking
place. Nevertheless, if this inversion symmetry is broken, for
example, by thermal poling,1 frequency doubling and the linear electro-optical effect can be induced. During the poling
process, a dc electric field (E dc) is recorded in the sample. In
thermal poling, E dc is believed to arise from a space-charge
distribution in the depletion layer near the anodic surface. An
(2)
effective second-order susceptibility ( ␹ eff
) as large as 1
1
pm/V can then be induced. Since the optical nonlinearity is
created in this layer or on its immediate vicinity, several
studies have been carried out to characterize this region of
the poled glass. The location, spatial extent, charge composition, and electric-field profile in the depletion layer depend
on the poling voltage, temperature, atmosphere, and poling
time.2–5 Some of the studies revealed that this region is a few
micrometers wide1,4–8 and that the width is larger for
vacuum poling than for poling in air.5 Further, in some cases
the depletion region was found to be neutral, and to finish at
a thin distribution of negative charge.5,6 Positive ion
in-diffusion3,7 and negative-charge emission3–5,9,10 have been
invoked to explain the neutrality of the layer. A correlation
was also found between the width of the depletion region
and the distribution of lithium ions in the sample.11
In this letter, we report on measurements of the
depletion-layer width and the average recorded electric field
in thermally poled silica as a function of the poling voltage.
The samples studied were 2-mm-thick, 20-mm-diam,
Vitreosil silica disks containing nominally 4 ppm Na, 4 ppm
Li, 2.5 ppm K, and 30 ppm Al. All samples were poled under
similar conditions except for the applied voltage, which was
varied from 2 to 8 kV. Poling was performed during 90 min
at T⫽280 °C, in air, with pressed-on electrodes. The diameter of the anode 共8 mm兲 was smaller than that of the silica
disks. In this way, the same sample had regions inside and
outside the anode contact, referred to as poled and reference
regions, respectively. All the poled samples generated
second-harmonic radiation upon excitation with Q-switched
and mode-locked Nd:YAG laser radiation at 1.064 ␮m. An
interferometric technique12 was used to accompany in real
time and with submicrometer precision the etching rate of
the poled samples. The anodic surface was etched with 20%diluted hydrofluoric acid while the intensity of the optical
interference between laser beams reflected by the cathodic
and anodic surfaces was recorded in time, for both reference
and poled regions.
Figure 1 shows a typical result of the interference-fringe
intensity as a function of the etching time for poled and
reference regions. One observes a constant oscillation period
T for the reference signal (T ref⫽60 s兲. For the poled signal
the period is longer during the first 1700 s (T pol⫽80 s兲 and
then becomes equal to the reference value T ref . It is known
that the etching rate of silica depends on the magnitude of
the electric field to which it is subjected,13 on the concentration of various doping cations,14,15 and on bond angles.16
However, even without identifying the reason behind the
a兲
Author to whom correspondence should be addressed; electronic mail:
triques@fis.puc-rio.br
b兲
Present address: Instituto de Fı́sica Gleb Wataghin, UNICAMP, 13083-970
Campinas-SP, Brazil.
c兲
Present address: ACREO, Isafjordsgatan 22, Electrum 236, Stockholm
16440, Sweden.
FIG. 1. Intensity of the interference fringes as a function of time for reference and poled regions. Sample poled under V⫽3 kV.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
0003-6951/2000/76(18)/2496/3/$17.00
2496
© 2000 American Institute of Physics
143.106.108.174 On: Fri, 26 Jun 2015 14:12:57
Triques et al.
Appl. Phys. Lett., Vol. 76, No. 18, 1 May 2000
2497
⫻108 V/m. The discrepancy can be attributed to the model,
which was tested only for weak fields. The average field
5.3⫻108 V/m is comparable to the 4.5⫻108 V/m figure calculated by Carlson17 under different poling conditions than
prevailing here. It is also in rough agreement with the breakdown field for the silica glass.18
If one attributes all the variation of the etching rate to the
recorded electric field,13 the above experimental findings imply that the field is directed from the anodic surface towards
the bulk and the electric-field profile across the depletion
layer is flat with amplitude E dc . The linear dependence of w
FIG. 2. Normalized etching rate of the reference region 共solid squares兲 and
with V is not in agreement with the simple one-ion model,
poled region for samples poled at V⫽2 kV 共triangles兲, 3.5 kV 共circles兲, and
6.5 kV 共open squares兲. The horizontal arrows indicate the depletion-layer
for which it is expected that the width depends on the square
width w for the sample poled at 3.5 kV.
root of the voltage.2,3,6 The linear dependence would be consistent with a constant electric field E dc , built up across a
width w and confined to it, with a magnitude independent of
variation in etching rate, it is still possible to identify the
the applied voltage, so that V⫽E dc⫻w. In this picture, due
depletion region as that where the etching rate is different
than in the bulk of the sample.12 Such a procedure was
to a multiple positive-charge carrier transport in the thermal
adopted in the present study. An additional feature seen in
poling process,5,7 a thin negatively charged layer would acFig. 1 is that the signal oscillation amplitude for the poled
cumulate at a distance w from the anodic surface, leaving a
region decreases temporarily during etching in the depletion
neutral depletion layer behind it. The neutrality of the deplelayer, as discussed later.
tion region would arise from positive-charge in-diffusion3,7
The etching rate of several samples poled under voltages
or negative-charge emission.3–5,9,10 Breakdown would set a
from 2 to 8 kV was measured and normalized to the average
limit to the maximum field in the heated glass to 5.3⫻108
rate of the reference region. Figure 2 shows results for three
V/m, limiting the value 共⫺␴兲 of the charged-layer density. A
samples poled at voltages 2, 3.5, and 6.5 kV. The normalized
positively charged layer 共⫹␴兲 would be attracted to the
etching rate ␳ is observed to be: 共i兲 constant across the deplesample surface due to the electric field, creating a system
tion layer; 共ii兲 smaller in the depletion layer than in the refsimilar to a plane–plate capacitor, with an electric field conerence: ␳ dl /␳ref⫽(0.67⫾0.04); and 共iii兲 independent of the
fined to the depletion region. This would give rise to a conpoling voltage. Experiments carried out by etching the castant E dc along w and independent of V.
thodic surface of a poled sample showed an etching rate
Assuming to a first approximation that the value of ␹ (3)
equal to that of the reference 共within the uncertainty of 5%兲.
of the glass 关 5⫻10⫺22 m2/V2 共Ref. 19兲兴 does not change
The depletion-layer width w was determined from the
much after poling,20 and considering that in the depletion
position at which the normalized etching rate for the poled
(2)
region ␹ eff
⫽3␹(3) E dc , another consequence of having the
region becomes equal to 1, as illustrated in Fig. 2 for the
electric-field value pinned to the average value 5.3⫻108 V/m
sample poled at 3.5 kV. Figure 3 shows a plot of w as a
is that the effective second-order nonlinear susceptibility crefunction of the applied voltage. The solid squares are the
ated would be constant across the depletion layer and equal
experimental data and the straight line is a linear fit to the
to 0.75 pm/V, irrespective of the poling voltage. Note, howdata. One observes that the thickness of the poled layer varever, that the simple model discussed above would not acies linearly with the poling voltage. The inverse of the slope
count for the reduction in the SH signal for increasing poling
represents the average electric field in the depletion region,
8
times, as reported in several studies.2,3,5
E dc⫽(5.3⫾0.3)⫻10 V/m. Consequently, the average reA couple more points can be noted. We believe that the
corded electric field is found to be independent of the poling
present
etching measurements are not in contradiction with
voltage. Its value is of the same order of magnitude of the
those performed by Alley and Brueck8 and Wong, Xu, and
electric-field mean value calculated using the present etching
Fleming,21 in which the etching rate of poled samples is
rate data and the model of Ref. 13, which gives (3.5⫾0.5)
modified only at the edge of the depletion region where the
fringing field creates a large component perpendicular to the
surface of the sample. Finally, as remarked above in regards
to Fig. 1, there is a strong loss of oscillation amplitude near
the inner edge of the depletion region. This loss of amplitude
can be attributed to nonuniformity in the etching rate due to
the proximity of the nonuniform distribution of negative
charges in the ⫺␴ layer.
In conclusion, we reported on measurements of the
depletion-layer width and the electric-field magnitude in
thermally poled fused silica. Under our poling conditions,
the depletion layer width w varies linearly with the poling
voltage: V⫽E dc⫻w. The average recorded field E dc⫽5.3
FIG. is
3. copyrighted
Width of theasdepletion
layer
as article.
a function
of the
poling
voltage.
This article
indicated
in the
Reuse
of AIP
content
is subject to 8the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
⫻10 V/m, is independent of the poling voltage.
Solid squares: experimental data; straight line: linear fit to the data.
143.106.108.174 On: Fri, 26 Jun 2015 14:12:57
2498
Two of the authors 共C.M.B.C. and V.B.兲 acknowledge
CNPq for their fellowships. The authors acknowledge
L. C. G. Valente for helpful discussions. This work was partially supported by PADCTIII 共FINEP-Brazil兲 and NEDO
International Joint Research Program 共Japan兲.
1
Triques et al.
Appl. Phys. Lett., Vol. 76, No. 18, 1 May 2000
R. A. Myers, N. Mukherjee, and S. R. J. Brueck, Opt. Lett. 16, 1732
共1991兲.
2
N. Mukherjee, R. A. Myers, and S. R. J. Brueck, J. Opt. Soc. Am. B 11,
665 共1994兲.
3
H. Takebe, P. G. Kazansy, P. St. J. Russel, and K. Morinaga, Opt. Lett.
21, 468 共1996兲.
4
P. G. Kazansky, A. R. Smith, P. St. J. Russell, G. M. Yang, and G. M.
Sessler, Appl. Phys. Lett. 68, 269 共1996兲.
5
V. Pruneri, F. Samoggia, G. Bonfrate, P. G. Kazansky, and G. M. Yang,
Appl. Phys. Lett. 74, 2423 共1999兲.
6
P. G. Kazansky and P. St. J. Russel, Opt. Commun. 110, 611 共1994兲.
7
T. G. Alley, S. R. J. Brueck, and R. A. Myers, J. Non-Cryst. Solids 242,
165 共1998兲.
8
T. G. Alley and S. R. J. Brueck, Opt. Lett. 23, 1170 共1998兲.
9
D. E. Carlson, K. W. Hang, and G. F. Sockdale, J. Am. Ceram. Soc. 55,
337 共1972兲.
10
U. K. Krieger and W. A. Lanford, J. Non-Cryst. Solids 102, 50 共1988兲.
11
T. G. Alley, S. R. J. Brueck, and M. Wiedenbeck, J. Appl. Phys. 86, 6634
共1999兲.
12
W. Margulis and F. Laurell, Opt. Lett. 21, 1789 共1996兲.
13
B. Lesche, F. C. Garcia, E. N. Hering, W. Margulis, I. C. S. Carvalho, and
F. Laurell, Phys. Rev. Lett. 78, 2172 共1997兲.
14
D. E. Carlson, K. W. Hang, and G. F. Sockdale, J. Am. Ceram. Soc. 57,
295 共1974兲.
15
C. M. B. Cordeiro, C. Borges, L. C. G. Valente, I. C. S. Carvalho, B.
Lesche, and W. Margulis, J. Non-Cryst. Solids 247, 183 共1999兲.
16
A. Agarwal and M. Tomozawa, J. Non-Cryst. Solids 209, 166 共1997兲.
17
D. E. Carlson, J. Am. Ceram. Soc. 57, 291 共1974兲.
18
W. T. Lynch, J. Appl. Phys. 43, 3274 共1972兲.
19
H. Imai, S. Horinouchi, N. Asakuma, K. Fukao, D. Matsuki, H. Hiroshima, and K. Sasaki, J. Appl. Phys. 84, 5415 共1998兲.
20
Note that a 1.8 times increase in ␹ (3) was reported by W. Xu, D. Wong,
and S. Fleming, Electron. Lett. 35, 922 共1999兲, for a thermally poled fiber.
21
D. Wong, W. Xu, and S. Fleming, in WDM Components, Optical Society
of America Trends in Optics and Photonics Series 共Optical Society of
America, Washington, DC, 1999兲.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
143.106.108.174 On: Fri, 26 Jun 2015 14:12:57