Kramers-Kronig analysis of soft x-ray reflectivity data of platinum thin film in 40-200 Å

wavelength region
Saurabh Sharma, R. K. Gupta, Mangalika Sinha, P. Yadav, Amol Singh, and Mohammed H. Modi

Citation: AIP Conference Proceedings 1731, 080001 (2016); doi: 10.1063/1.4947879

View online: http://dx.doi.org/10.1063/1.4947879
View Table of Contents: http://scitation.aip.org/content/aip/proceeding/aipcp/1731?ver=pdfcov
Published by the AIP Publishing

Articles you may be interested in

Magneto-optical Kramers-Kronig analysis
Rev. Sci. Instrum. 86, 033906 (2015); 10.1063/1.4914846

Kramers‐Kronig Analysis of Infrared Reflectance Spectra for Quaternary In x Al y Ga 1−x−y N Alloy

AIP Conf. Proc. 1250, 337 (2010); 10.1063/1.3469673

KKRAMERS: Kramers‐Kronig Analysis using Maximum Entropy for Reflectance Spectroscopy

AIP Conf. Proc. 1193, 413 (2009); 10.1063/1.3275643

Kramers–Kronig analysis of polymer acoustic data

J. Appl. Phys. 85, 44 (1999); 10.1063/1.369406

Sum‐rule constraints in reflectance extrapolation for Kramers‐Kronig analysis

J. Appl. Phys. 46, 3066 (1975); 10.1063/1.321999

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions IP: 140.110.200.36 On: Tue, 21 Jun 2016 09:42:06
 Kramers-Kronig Analysis of Soft X-ray Reflectivity Data
of Platinum Thin Film in 40-200 Å Wavelength Region
Saurabh Sharma1, 2, R.K. Gupta1, Mangalika Sinha3, P. Yadav1,3, Amol Singh3,
Mohammed H. Modi1,3*
1
Indus Synchrotrons Utilization Division, Raja Ramanna Centre for Advanced Technology, Indore 452 013,
India
2
Department of Applied Physics & Opto-Electronics, Shri Govindram Seksaria Institute of Technology and
Science, Indore 452 003, India
3
Homi Bhabha National Institute, Raja Ramanna Centre for Advanced Technology, Indore 452 013, India
*E-mail: modimh@rrcat.gov.in

Abstract. Reflectivity beamline at Indus-1 synchrotron source is used to determine optical constants of a platinum
thin film in the soft x-ray wavelength region of 40-200Å by applying Kramers-Kronig (KK) technique on R vs
wavelength data. Upto 150Å wavelength region the results of KK analysis are found in good agreement with the
Henke’s optical constants and also with those obtained by the angle dependent reflectivity technique. A significant
mismatch is observed above 150Å wavelength region which could be due to the presence of higher harmonics in the
toroidal grating spectra of the reflectivity beamline.
Key words: Soft x-ray, Optical constants, X-ray reflectivity, Multilayers, Beam line.
PACS:78.20.Ci; 41.60.Ap; 41.50.+h

INTRODUCTION region and therefore different interpolation/

extrapolation techniques are used.
In the x-ray region, the optical properties of any In this paper, the optical constants of a platinum
material are defined by energy dependent complex thin film (450Ǻ thickness) are determined in the 40-
refractive index n = (1 - δ) + iβ, where δ and β are the 200 Ǻ wavelength region using the angle dependent
optical constants which represent the dispersive and reflectivity technique and the Kramers-Kronig relation.
absorptive behavior of the material. A data base of real In order to apply the Kramers-Kronig method, the
and imaginary part of refractive index, known as reflectivity data over large spectral range is required.
“1993 atomic table”1, has been published in 1993. An Therefore, outside the measurement region, the
updated version of this database is available at the reflectivity values taken from theoretical calculations
CXRO website 2 . 7he measurements of the optical are used for KK analyses.
constants are subjected to a number of difficulties
inherent to the soft x-ray/ vacuum ultra-violet region. THEORETICAL BACKGROUND
Significant mismatch between the experimental values
and the tabulated values of the optical constants are The reflectivity of an air to film interface can be
found by different workers for many materials. calculated using the Fresnel equation4. For s-polarized
The experimental values of the optical constants can geometry this equation can be written as4
2
be determined using different techniques. In the angle 2
sin T  n 2  cos 2 T
R r
dependent reflectivity technique, the real and 2
sin T  n 2  cos 2 T
imaginary part of the optical constants are
simultaneously determined at discrete photon energies. Here, n, θ, R, r are the complex refractive index,
Using the Kramers-Kronig relation 3 , one can obtain grazing incident angle, reflectivity, and reflection
optical constants over large wavelength region by coefficient respectively.
using either the absorption data or the reflection data. The Fresnel reflection coefficient is a complex
There are inherent limitations associated with the quantity. The amplitude and phase relation can be
Kramers-Kronig relation as the integration range is given by the relation
defined as the zero to infinite energy region.
Ppractically it is not possible to obtain either r R exp(i\ )
absorption or reflection data for the zero to infinity

DAE Solid State Physics Symposium 2015

AIP Conf. Proc. 1731, 080001-1–080001-3; doi: 10.1063/1.4947879
Published by AIP Publishing. 978-0-7354-1378-8/$30.00

080001-1
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions IP: 140.110.200.36 On: Tue, 21 Jun 2016 09:42:06
 The phase information can be obtained by applying layer are found to be 17.3Ǻ, 501Ǻ, and 17.3Ǻ
Kramers-Kronig integral relation: respectively. The roughnesses of these three layers are
found to be 1.5Ǻ, 6.5Ǻ, 4.9Ǻ respectively. The
Z f ln R Z ' roughness of the silicon substrate is found to be 5.4Ǻ.
S ³0 Z ' 2 Z 2
\ Z  pv dZ '

where pν stands for the principal value. 3 Pt layer

Substrat region
rho u 10 (Ë )
In this formalism, the phase \ is determined by

-2

Surface region
0
10

Oxide layer
2
the KK analysis of the measured reflectivity data. The

-4
value of R is known from the experimental data. The -1
1
10
phase spectrum \ Z obtained by the KK analysis of

Reflectivity
0
the complex reflection coefficient r Z contains an 10
-2 0 200 400 600
DepthOËGP
inherent error of ±S 5. In the present study this error is
considered in the phase calculations. 10
-3

EXPERIMENTAL DETAILS 10
-4 Experimental data
Fitted data
The soft x-ray reflectometer installed at the
0.0 0.5 1.0 1.5 2.0
reflectivity beamline6 at Indus-1 synchrotron radiation Incident angle (deg.)
source was used for the angle dependent reflectivity
and the wavelength dependent reflectivity FIGURE 1. Measured and fitted GIXRR spectra of Pt thin
measurements. These measurements were carried out film obtained using O =1.54Ǻ are shown. In the inset, the
in the 40 to 200 Å wavelength region. SLD profile obtained from the fit parameters is shown.
For the study of soft x-ray optical constants of Pt
thin film by applying the KK analysis technique, a thin The measured and fitted soft x-ray reflectivity
film of ~ 450 Å thickness was deposited on a silicon (SXR) data in the 60 to 200Ǻ (in 20Å step)
substrate. The grazing incident x-ray reflectivity wavelength range are shown in the Fig. 2. The SXR
(GIXRR) measurements were performed using Bruker data points are fitted with the same three layer model
D-8 system which consists of a goniometer and a as used in the GIXRR analysis. The thickness and
copper Kα x-ray source (1.54 Å). For the reflectivity roughness values obtained from the GIXRR analysis
analysis, the Parratt formalism 7 was used and the are kept fixed for optical constants determination using
effect of the surface roughness was taken into account the SXR technique. This approach reduces the number
using the Névot-Croce model8. of fitting parameters and leaves the optical constants
as the only variables during the fit process. The results
RESULTS AND DISCUSSIONS of the optical constants obtained are given in the
Tables 1 and 2, along with the Henke’s values.
The soft x-ray measurements on the Pt thin film are
carried out for determining the optical constants. Prior In the angle dependent reflectivity method, the
to that the grazing incidence x-ray reflectivity optical constants are obtained in a discrete interval of
(GIXRR) technique was used to determine the film’s 20Å in the wavelength region of 40-200Å. In order to
structural parameters i.e. film thickness, interface obtain a continuous values of the optical constants, the
roughness, and film density. The measured and fitted Kramers-Kronig method is used. For the KK analysis,
GIXRR curves for the Pt thin film are shown in the the wavelength dependent reflectivity data are
Fig. 1. The best fit curve is obtained using a 3 layer measured for 3 degree incidence angle. To perform
model. The model comprised of an oxide layer on a Si KK integration, it is required to have the reflectivity
substrate, a principal Pt layer, and a top contamination data over a large wavelength range. Therefore, in the
layer which generally forms due to the interaction of present case, the reflectivity values obtained from the
the surface with the ambient. The scattering length theoretical calculations are used to increase the data
density (SLD) profile obtained from the GIXRR fit is range beyond the experimental measurement region.
shown in the inset of the figure. The SLD profile
suggests that the film is uniform across the depth
except at the air/film interface and at the film/substrate
interface. The density obtained from the GIXRR
analysis is found to be 20.8 g/cm3, which is 97% of the
Pt bulk density (21.4g/cm3). The thicknesses of the
oxide layer, the principal layer, and the contamination

080001-2
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions IP: 140.110.200.36 On: Tue, 21 Jun 2016 09:42:06
 measured data
from the KK analysis and those obtained from the
fitted data angle dependent reflectivity measurements are plotted
200‡ in the Fig. 3 along with the Henke’s values. It is
evident that the optical constants values obtained by
Intensity (arb units)

180‡
the KK analysis and by the R vs T technique are in a
160‡
good agreement upto 150Å wavelength region. Above
140‡
the 150Å wavelength region the large difference is
120‡
observed, which could be due to the presence of higher
100‡ harmonics contamination in the incident beam
80‡ obtained from the toroidal grating monochromator in
60‡
the reflectivity beamline.
0.15
KK analyses
0 10 20 30 40 50 60 70
angle dependent
Incident angle(degree) 0.10 reflectivity method
Henke data
G
FIGURE 2. SXR spectra of Pt thin film recorded in the 60- 0.05

200Ǻ wavelength region in 20Å step are shown. For the sake
0.00
of clarity, the curves are vertically shifted.
0.3 Henke data
angle dependent
TABLE 1. Comparison of the measured delta values 0.2
reflectivity
KK analyses
obtained from the angle dependent reflectivity technique E
and the Kramers-Kronig method with the Henke’s data 0.1

Wavelength / Delta 0.0

Energy
R vs T R vs O by Henke 0 50 100 150 200 250
(Ǻ / eV) values
method KK method Wavelength (Ë)
60 (206.67) 0.01378 0.012 0.01326 FIGURE 3. Comparison of the optical constants obtained by
80 (155) 0.0255 0.024 0.02363 the KK analysis with the Henke tabulated values and those
100 (124) 0.0499 0.045 0.0487 obtained from aangle dependent reflectivity experiments.
120 (103.33) 0.0859 0.083 0.08099
140 (88.57) 0.11 0.10 0.116
160 (77.5) 0.1165 0.11 0.1245 CONCLUSIONS
180 (68.88) 0.104 0.124 0.0841
200 (62) 0.1382 0.12 0.0975 The optical constants of platinum are obtained
using the Kramers-Kronig relation and the angle
TABLE 2. Comparison of the measured beta values dependent reflectivity measurement in the 40-200Ǻ
obtained from the angle dependent reflectivity technique wavelength region. The optical constants obtained
and the Kramers-Kronig analysis with the Henke’s data using the Kramers-Kronig relation and the angle
dependent reflectivity measurement are similar for the
Wavelength / Beta
Energy
40-150Ǻ wavelength region. Above 150Ǻ
(Ǻ / eV) R vs T R vs O by Henke disagreement occurs between the optical constants
method KK method values because of the higher harmonic contamination, which
60 (206.67) 0.015 0.010 0.0146 affects the measured reflectivity data.
80 (155) 0.0107 0.010 0.01358
100 (124) 0.01476 0.014 0.0173 REFERENCES
120 (103.33) 0.03 0.042 0.03504
140 (88.57) 0.05284 0.077 0.0738
160 (77.5) 0.0796 0.12 0.1346 1
B.L. Henke et al, At. Data Nucl. Data Tables 54, 181-342,
180 (68.88) 0.1181 0.15 0.1621
1993.
200 (62) 0.1338 0.182 0.1712 2
www.cxro.lbl.gov
3
E. Filatova et al, Appl. Opt. 49, 2539-2546, 2010.
The measured R vs O data in 40-225Å range are 4
D. Attwood, Cambridge University Press, 1999.
merged with the calculated reflectivity data to generate 5
P. L. Nash et al, J. Mod. Opt. 42, 1837-1842, 1995.
the reflectivity vs wavelength data over a large spectral 6
region of 0.5 to 300Ǻ. This extended data are used to R. V. Nandedkar et al, Curr. Sci. 82, 3, 2002.
7
obtain the phase information by using the KK relation L. G. Parratt, Phys. Rev. 95, 359-369, 1954.
8
that in turn are used to determine the delta and beta L. Névot and P. Croce, Rev. Phys. Appl. (Paris) 15, 761–
values. The values of the optical constants obtained 779, 1980.

080001-3
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions IP: 140.110.200.36 On: Tue, 21 Jun 2016 09:42:06