CHIN.PHYS.LETT.
Vol. 25, No. 4 (2008) 1321
Influence of Different Substrates on Laser Induced Damage Thresholds at
1064 nm of Ta2 O5 Films ∗
XU Cheng(许程)1,2∗∗ , MA Jian-Yong(麻健勇)1,2 , JIN Yun-Xia(晋云霞)1 , HE Hong-Bo(贺洪波)1 ,
SHAO Jian-Da(邵建达)1 , FAN Zheng-Xiu(范正修)1
1
Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, Shanghai 201800
2
Graduate School of Chinese Academy of Sciences, Beijing 100049
(Received 16 January 2008)
Ta2 O5 films are prepared on Si, BK7, fused silica, antireflection (AR) and high reflector (HR) substrates by
electron beam evaporation method, respectively. Both the optical property and laser induced damage thresholds
(LIDTs) at 1064 nm of Ta2 O5 films on different substrates are investigated before and after annealing at 673 K for
12 h. It is shown that annealing increases the refractive index and decreases the extinction index, and improves
the O/Ta ratio of the Ta2 O5 films from 2.42 to 2.50. Moreover, the results show that the LIDTs of the Ta2 O5
films are mainly correlated with three parameters: substrate property, substoichiometry defect in the films and
impurity defect at the interface between the substrate and the films. Details of the laser induced damage models
in different cases are discussed.
PACS: 42. 79. −e, 68. 60. −p, 81. 15. Ef
The laser induced damage thresholds (LIDTs) of
optical films have become more important with the increasing high average output and high peak power of
lasers.[1] The mechanism of laser induced damage has
not been fully understood as it is complex and associated with many factors. The initiation of the damage of films varies with deposition methods and conditions, post-treatment and laser characteristics.[2−4]
For the nanosecond laser pulse, the damage of films
has always been induced by defects, and the LIDTs
can be improved by decreasing the density or enhancing the mechanical stability of defects.[5,6] The
detailed damage mechanism and models are different
and should be discussed according to different cases.
Previous reports showed that the increase of deposition rates may increase the LIDTs by decreasing
the void defect in films, whereas annealing increases
the LIDTs by improving the substoichiometry defect,
and SiO2 protective layers are effective to improve the
laser damage resistance by improving heat conduction
around the inner defect of Ta2 O5 films.[7,8] To our
best knowledge, though much attention have been extensively paid to the LIDTs of films, little has been
studied about the effect of different substrates on the
LIDTs of single layer films. In this study, we investigate the influence of five substrates, which contains
Si, BK7, fused silica, antireflection (AR) and high reflector (HR) substrates, on the LIDTs of Ta2 O5 films
before and after annealing.
In our experiment, Ta2 O5 films with an optical
thickness of 6 QWOT (quarter wavelength optical
thickness) at a wavelength of 550 nm were deposited
on Si, BK7, fused silica, antireflection (AR) and high
reflector (HR) substrates. Designs of AR and HR sub∗ Supported
strates are Air|0.5H1.25L|glass and air|(HL)10 |glass,
respectively, where H stands for QWOT of ZrO2 and
L stands for QWOT of SiO2 (the referent wavelength
λ is 1064 nm), and glass is BK7. The transmittance
of AR substrate at 1064 nm is 99.52% and the reflectance of HR substrate is 99.68%, respectively. The
films were deposited when the chamber was pumped to
2×10−3 Pa and oxygen was introduced to keep oxygen
partial pressure of 2 × 10−2 Pa. Annealing of the films
was performed in air at 673 K for 12 h. Transmittance
T and reflectance R at 1064 nm of the substrates was
measured by a Lambda 900 spectrophotometer and
the measurement error was within 0.08%. The absorption A was calculated by the equation (scattering
is neglected)
A = 1 − T − R.
Transmittance spectra of the deposited Ta2 O5
films were also measured by the Lambda 900 spectrophotometer. The optical constants were obtained
by photometric method from the measured transmittance spectral curves.[9] The composition of the films
were analysed by XPS using focused (300 µm in diameter) monochromatic Al-Kα (hν = 1486.6 eV) radiation at a pass energy of 20 eV. The oxygen to tantalum
ratio in the films was determined by peak deconvolution of the XPS curves. Damage testing was performed in the 1-on-1 regime, using 1064 nm Q-switch
pulsed laser at a pulse length of 12 ns. The experimental setup is shown in Ref. [10]. The damage morphologies of samples were observed by a Leica DMRXE micropolariscope.
The absorption and LIDTs of the substrates are
listed in Table 1. It is found that the absorption
by the National Natural Science Foundation of China under Grant No 60608020.
Email: [email protected]
c 2008 Chinese Physical Society and IOP Publishing Ltd
°
∗∗
(1)
1322
XU Cheng et al.
of Si substrate is 70.32%, which is much higher than
that of other substrates. The LIDT of Si substrate is
2.1 J/cm2 , and is the least in all the substrates. This
is mainly due to the strong absorption which is more
easily to adsorb laser energy and induced damage.
Among the five substrates, the fused silica achieves
the highest LIDT.
Table 1. Absorption and LIDTs of the substrates.
Substrates
Si
BK7
fused silica
AR
HR
Absorption A
70.32%
< 0.08%
< 0.08%
0.11%
0.12%
LIDT (J/cm2 )
2.1
32.7
36.8
15.2
30.5
Vol. 25
tive index increase and the extinction index decrease
after annealing. It indicates that annealing is beneficial to improve the film density and to decrease the
optical loss. Thus, the adhesive force of the film increases and the film absorption decreases, which is
effective to improve the laser resistance.
All the films before and after annealing were examined by XPS in order to obtain the film chemical
structure. The high-resolution spectra are similar for
all the five samples. Figure 3 illustrates the Ta 4f
spectra of the typical sample before and after annealing. It reveals two 4f7/2 and 4f7/2 peaks at 26.4 and
28.2 eV, respectively, which shows composition of the
film is Ta5+ .[11] The O/Ta ratio is 2.42 for the asdeposited film and 2.50 for the annealed film. The
improvement of the stoichiometry of the Ta2 O5 film
is beneficial to the decrease of the film absorbance.
Fig. 1. Transmittance curves of the films.
Fig. 3. XPS results of the typical sample.
Fig. 2. Refractive index n and extinction index k of the
Ta2 O5 film on the BK7 substrate.
Figure 1 shows the transmittance curves of the
films on different substrates. It reveals that the transmittances of Ta2 O5 films at 1064 nm on different substrates are different. The Ta2 O5 film on a BK7 substrate was choose as the typical sample to study the
difference on the refractive index n and extinction index k before and after annealing. Figure 2 shows the
characteristics of n and k of the Ta2 O5 film on the
BK7 substrate. It can be clearly seen that the refrac-
The LIDTs of the samples before and after annealing are shown in Fig. 4. It is found from Fig. 4(a)
that before annealing, the LIDT of the film on the Si
substrate is 1.7 J/cm2 and is the lowest. The LIDTs
of other films are almost the same, about 4.4 J/cm2 ,
which have no relation with the LIDTs of substrates.
After annealing, all the LIDTs of samples increase.
The LIDT of the film on the Si substrate increases
slightly, while the LIDTs of other films increase significantly, about 50–90%, compared to the as-deposited
film. The improvement of the LIDT after annealing is
mainly due to the decrease of the substoichiometry defect in the film. The defect is mainly originated from
oxygen vacancy and the activation energy is lower
than 0.6 eV, which is much smaller than the band gap
of Ta2 O5 (4.3 eV).[14,15] Therefore, the defect in the
film is liable to adsorb more laser energy and easily
turns into the damage centre under laser radiation.
In addition, the decrease of voids in the film benefits
to the film density and the adhesive force among the
atoms, which also improves the LIDT. Furthermore, it
No. 4
XU Cheng et al.
is interesting to find from Fig. 4(b) that the LIDTs of
the annealed films have some relation with the LIDTs
of substrates. It is revealed that the higher the LIDT
of the substrate is, the higher the LIDT of the film is.
Fig. 4. LIDT results of the samples before (a) and after
(b) annealing.
Figure 5 shows the damage morphologies of the
five as-deposited and the typical annealed samples irradiated at different laser energies. There is some
difference in the damage morphologies between the
film on the Si substrate and the others. Figure 5(a)
shows a total fused damage area which is due to the
melt of the substrate. Figures 5(b)–5(e) show that the
damaged sites are centred on one or more absorption
points, which is attributed to the existence of defect.
In addition, the damage morphologies of the films after annealing are similar to themselves before annealing. The annealed film deposited on BK7 is chosen as
the typical sample as shown in Fig. 5(f). A little difference can be found except that the as-deposited sample
has a pit with a larger radius than the annealing sample, which implies that the film has an improved laser
resistance after annealing. It is indicated that annealing has little effect on the damage morphologies.
According to the LIDT results and damage mor-
1323
phologies, the effect on the damage of films can be
attributed to three parameters: the substrate property, substoichiometry defect in the film and impurity
defect at the interface between the substrate and the
film.[7,8,16] However, only one or two of them can dominate the LIDT in certain case. In our experiment,
three damage models dominated by different parameters are proposed as follows:
(1) Substrate property dominant model. It is
worth noting that in the five substrates used in our
experiment, the Si substrate is different from the other
four substrates. It has a strong absorption up to
70.32%, much higher than the others. The LIDT of
the Si substrate is merely 2.1 J/cm2 . The Si substrate
is easily melted and damaged under laser radiation
attributed to the strong absorption. In this case, the
LIDTs of films are dominated by the substrate property especially the laser damage resistance. The LIDT
of the Ta2 O5 film on the substrate is only 1.7 J/cm2
before annealing and slightly increases to 1.8 J/cm2
after annealing.
(2) Substoichiometry defect dominant model. It
is found that the other four substrates have less absorption and much higher LIDTs than those of the Si
substrate (Table 1). In this case, the substoichiometry defect in the films starts to dominate the damage
initiation. The density of this defect can be depicted
by the deviation of the theoretical O/Ta ratio. The
more deviation of the theoretical O/Ta ratio will induce more substoichiometry defect density and lower
LIDT.[16,17] As the five kinds of Ta2 O5 films are deposited by the same method and under the same conditions, the density of substoichiometry defect in the
films is also the same. It corresponds to the results
that the LIDTs are almost the same for the films on
different substrates except the Si substrate.
(3) Impurity defect and substrate property combined dominant model. After annealing, the O/Ta
ratio is enhanced to 2.50 as shown in XPS results,
and the substoichiometry defect disappears. The impurity defect at the interface between the substrate
and the film then becomes an important factor in the
laser damage course. At the meanwhile, as the damage initiates at the interface between the substrate
and the film, the substrate property will also affect
the damage of films. Thus, the combined two parameters dominate the LIDTs of films. Since the density
of impurity defect at the interface is about the same
due to the same method for cleaning the substrates
before film deposition, its effect on the LIDT is equal
and can be neglected. Therefore, the LIDTs of the
films are mainly affected by the difference of substrate
property. The results show that after annealing, the
higher the LIDT of the substrate is, the higher the
LIDT of the film is.
1324
XU Cheng et al.
Vol. 25
Fig. 5. Damage morphologies of the as-deposited samples on (a) Si, (b) BK7, (c) fused silica, (d) AR, (e) HR and the
annealed sample on (f) BK7.
In summary, the influence of different substrates
on the LIDTs at 1064 nm of Ta2 O5 films before and
after annealing has been studied. The LIDT of the
Ta2 O5 film on the Si substrate is much less than the
films on other substrates, which shows that different
substrates play different roles in the LIDTs. Moreover, before and after annealing, the LIDTs of Ta2 O5
films on the same substrates have different relations
with the LIDTs of substrates. This implies that the
same substrate may play different roles in the LIDTs
in different cases. The damage of the films is related
to three parameters in our study, and different damage models are established according to the dominant
parameters. The results will be beneficial to better
knowing the laser damaged mechanism of films.
References
[1] Gao H L and Wang N Y 1993 Appl. Opt. 32 7084
[2] Yu H, Qi H J, Cui Y, Shen Y M, Shao J D and Fan Z X
2007 Appl. Surf. Sci. 253 6113
[3] Zhao Y A, Wang Y J, Gong H, Shao J D and Fan Z X 2003
Appl. Surf. Sci. 210 353
[4] Yuan L, Zhao Y A, Shang G Q, Wang C R, He H B, Shao
J D and Fan Z X 2007 J. Opt. Soc. Am. B 24 538
[5] Zhang D P, Shao J D, Zhang D W, Fan S H, Tan T Y and
Fan Z X 2004 Opt. Lett. 29 2870
[6] Zhao Y A, Shao J D, He H B and Fan Z X 2005 Proc. SPIE
5991 599117-1
[7] Xu C, Dong H C, Ma J Y, Jin Y X, Shao J D and Fan Z X
Chin. Opt. Lett. Accepted
[8] Xu C, Li D W, Ma J Y, Jin Y X, Shao J D and Fan Z X
2008 Opt. Laser Technol. 40 545
[9] Manifacier J C, Gasiot J and Fillard J P 1976 J. Phys. E:
Sci. Instrum. 9 1002
[10] Hu H, Fan Z and Luo F 2001 Appl. Opt. 40 1950
[11] Szymanowski H, Zabeida O, Klemberg-Sapieha J E and
Martinu L 2005 J. Vac. Sci. Technol. A 23 241
[12] Kim I L, Kim J S, Kwon O S, Ahn S T, Chun J S and Lee
W J 1995 J. Electron. Mater. 24 1435
[13] Pignolet A, Rao G Mand Krupanidhi S B 1995 Thin Solid
Films 258 230
[14] Sawada H and Kawakami K 1999 J. Appl. Phys. 86 956
[15] Demiryont H, Sites J R and Geib K 1985 Appl. Opt. 24
490
[16] Yao J K, Shao J D, He H B and Fan Z X 2007 Appl. Surf.
Sci. 253 8911
[17] Yao J K, Fan Z X, Jin Y X, Zhao Y A, He H B and Shao
J D 2008 Thin Solid Films 516 1237