-
38th International Conference of IMAPS-CPMT Poland
Rzeszw - Czarna, 21-24 September 2014
1
Composition analysis of epitaxial NbTiN films for superconductor
photon detectors
Sylwia Przywska 1, Marek Guziewicz 2, Marcin Juchniewicz 2,
Renata Kruszka 2,
Edyta Piskorska-Hommel 3, Jarosaw Domagaa 4, Alexandru Marin
5,
Petre Osiceanu5, Andriej Klimov 2, Jan Bar2, Wojciech Sysz2
,
1 Faculty of Physics, Warsaw University of Technology, ul.
Koszykowa 75, 00-662 Warsaw, Poland;
2Institute of Electron Technology, al. Lotnikow 32/46, 02-668
Warsaw, Poland; 3 Institute of Solid State Physics, University of
Bremen, Otto-Hahn-Allee, 28359 Bremen,
4 Institute of Physics, PAS, al. Lotnikow 32/46, 02-668 Warsaw,
Poland 5 Institute of Physical Chemistry, Spl. Independentei 202,
Bucharest, Romania
Abstract: We report the characterization of the ultrathin NbTiN
films for SSPDs. The higher quality of the ultrathin
superconducting films in comparison to niobium nitride, as far as
the fabrication technology of single photon detectors is concerned,
was demonstrated. The films deposited on Al2O3 single crystals
shown excellent both superconducting and structure properties. New
results based on XPS studies of NbTiN films reveal presence of some
contaminations like carbon and oxygen. The following XPS peaks were
examined: Nb 3d, Ti 2p, O 1s, N 1s, C 1s and Al 2p. Compounds of
NbN, NbTiN and some Nb-oxides have been revealed. The NbTiN films
with thickness of 4 nm, grown on the Al2O3 and post-grown annealed
at 1000
oC in Ar, reach critical temperature of 14K. Moreover, the films
disclose the best superconducting properties - extremely high
critical current density of 12106 A/cm2.
Key words: Superconductor, NbN, XPS, single-photon detector
1. INTRODUCTION
Superconducting single-photon detectors (SSPDs) are able to
detect single optical photons, these have relatively high quantum
efficiency and low dark counts rate and low jitter. The detectors
are expected to play a leading role in such applications as optical
quantum information processing, satellite communications and
medical diagnostics, especially as detectors of singlet oxygen
luminescence in photodynamic therapy. Construction of a detector
includes 100 nm wide stripes patterned in an ultrathin NbN or NbTiN
film. They are biased on a subcritical current. The absorption of a
photon generates a hot spot that grows until a resistive region is
formed across the nano-stripe, thus, produced a detectable voltage
pulse. The higher quality of ultrathin superconducting NbTiN films
in comparison to NbN films was demonstrated. High epitaxial quality
of NbTiN films grown on the Al2O3 substrates was proved by HRXRD in
our previous work [1,2], but film composition was not cleared
because of problems regarding to composition study on so ultrathin
films. New results for NbTiN and NbN films that based on X-ray
photoelectron spectroscopy (XPS) studies reveal presence of
contaminations such as carbon and oxygen.The quantitative
composition analyses of NbTiN films as well as detection parameters
of SSPD made with the NbTiN film are here presented.
2. TECHNOLOGY OF SUPERCONDUCTIONG FILMS
The NbTiN films were grown by high-temperature reactive
radio-frequency magnetron sputtering using 1000C system from the
Surrey Nanosystems Ltd. The films were deposited from a 3-inch
diameter Nb and Ti targets at DC power of 220 W on Nb and of 80 W
on Ti, respectively, in N2Ar plasma at temperature of 850oC.
Al2O3(0001) and Si(001) are used as substrates. The thicknesses
of
-
38th International Conference of IMAPS-CPMT Poland
Rzeszw - Czarna, 21-24 September 2014
2
the studied layers was 4 nm. To improve superconducting
properties of the films Rapid Thermal Annealing (RTA) was conducted
at 1000oC in Ar for 10 min [1].
3. XPS STUDIES
3.1. Quantitative analysis of film composition XPS
investigations of NbTiN films deposited on Si and Al2O3 were
performed on samples assigned NbTiN 11Si and NbTiN 11sh,
respectively. The XPS spectra were recorded by means PHI Quantera
SXM which is the state of the art Scanning XPS Microprobe equipment
in Bucharest. The studied films were cleaned in-situ by sputtering
using Ar+ at 500 eV. Analysis of XPS spectra concerns following
peaks: Nb 3d, Ti 2p, O 1s, N 1s, C 1s and Al 2p or Si 2p. The XPS
quantification is performed by assigning quantification regions,
subtracting the background from each region. We used two methods of
background subtracting: Spline Shirley (used for titanium peaks),
which requires setting up several nodes, necessary for the Shirley
algorithm, and Shirley (used for other component peaks). Percentage
atomic concentration parameter is computed from the raw peak area
divided by the Relative Sensitivity Factor (RSF), which is
extracted from the given library for every peak identified by a
region. The atomic concentration Xi is computed using the
formula:
==
m
i i
ii
A
AX
1
100 (1)
where Ai (i = 1, 2, ) are the adjusted intensities, which are
determined from the measured intensity Ii, the transmission
function Ti evaluated for electrons of recorded energy E, the
relative sensitivity factor Ri for the transition i and the escape
depth compensation exponent n. The adjusted intensities are defined
by eq. 2 as follows:
( ) ini
i REET
IA 100= (2).
Multiplying Xi by the atomic mass of the components and
normalising to 100% we can get mass % concentration. For
quantitative analysis of Nb 3d and Ti 2p peaks we used the
procedure given in papers [3, 4]. The intensity ratio of doublets
2p3/2 and 2p1/2 peaks are constrained to be at a ratio of 2:1, the
intensity ratio of 3d3/2 and 3d5/2 doublets are constrained to be
at a ratio of 2:3. FWHM of the peaks in doublets are forced to be
alike, except for the Ti doublets in TiO2,because of Coster-Kroning
effect, where FWHM of the 2p1/2 peak is broader than FWHM of the
2p3/2 peak.
The values of binding energy applied in our study are the
average Binding Energy (BE) data from the NIST XPS database, or the
values taken from the paper [3]. The binding energies, which are
used in this work are shown in the table 1. The doublet splitting
was constrained as we used the available literature data.
Tab. 1. Binding energies E(eV) of Nb and Ti components [3]
applied in our XPS spectra simulations.
Peak NbN Nb2O5 NbN(1-x)O(x) NbNO/NbCO Nb2N(2-y)O(3-y) TiO TiN
Ti2O3 TiO2
3d5/2 203.7 207.5 204.8 205.9 206.9 - - - -
3d3/2 206.5 209.9 206.8 208.8 209.5 - - - -
2p1/2 - - - - - 460.9 461.8 462.3 464.3
3.2. Composition of NbTiN films Samples NbTiN11sh and NbTiN11Si
were taken under the investigation. Based on full XPS spectrum and
using the formula (1) we computed atomic % concentrations for the
samples and included in Tab.2. The concentrations of components
change slowly with the sputtering time. As expected the
concentration of carbon suddenly decreases after the first
sputtering step, then the falling trend maintains slowly.
Concentration of oxygen is gradually decreasing with the etching
time, while Al
-
38th International Conference of IMAPS-CPMT Poland
Rzeszw - Czarna, 21-24 September 2014
3
concentration increases slowly with the time, so we can suppose
that oxygen is bonded with metals, more intensely on surface, while
the Al signal relates to the sapphire substrate. Analyzing the Nb3d
peak shown in Fig.1a., there are visible doublets 3d5/2, and 3d3/2
of the following components: NbN, Nb2O5, NbN1-xOx, Nb2N2-yO3-y, and
NbNO with NbCO (which are not bonded together, but differences in
their binding energy are too small to subtract their spectra). In
the Ti2p peak there are visible doublets 2p3/2, and 2p1/2 which can
be assigned to the following components: TiO, TiN, Ti2O3 and TiO2.
The relative ratio of the noted components changes with the
sputtering time. The obtained data for distribution of mass conc.
of Nb and Ti compounds is presented in Tab. 3.
Tab. 2. At.% composition of the NbTiN11sh film before and after
sputtering etching of the surface. Sputtering time Nb Ti O N C
Al
0 25.0 1.8 20.9 17.2 33.2 1.9
0.1 min 39.1 2.0 21.5 26.4 8.5 2.4
0.3 min 43.7 2.4 17.2 28.6 5.1 2.8
0.5 min 45.5 3.0 14.0 30.00 3.3 3.9
a) b)
Fig. 1. An example of Nb3d (a) and Ti2p (b) peaks from XPS
spectra measured on the NbTiN 11sh sample as-
deposited on sapphire substrate.
Tab. 3. Distribution of mass % concentration of compounds
identified in the NbTiN 11sh film as-dep. Sputtering time
NbN Nb2O5 NbN(1-x)Ox NbNO/NbCO Nb2N(2-y) O(3-y) TiO TiN Ti2O
3
TiO2
As rec 21.7 15.5 34.6 10.0 18.1 17.5 11.8 41.9 28.8
0,2 min 27.9 20.4 43.5 8.2 - 16.6 30.1 30.5 22.8
0,5 min 33.1 17.3 47.1 2.5 - 7.2 29.6 28.9 34.3
Although, as expected, the NbN atomic concentration level is
very high, our data shows that there is substantial amount of
NbN(1-x)Ox. The concentration levels of NbN as well as NbN(1-x)Ox
is increasing with the sputtering time. The concentration of Nb2O5
is mostly visible at the surface of the sample, then its level
decreases with the time. The NbNO/NbCO concentration ratio, which
indicates the level of contamination of our sample, is decreasing
with the sputtering time. Undoubtedly, Nb2N (2-y)O(3-y) is present
only on the surface of the sample. The titanium components on the
surface are mostly oxides, but the TiN mass concentration increases
with the sputtering time. In the case of NbTiN film annealed at
1000oC (NTN11shW) we calculated based on XPS measurements similar
at. % concentration of Nb and Ti, but oxygen contamination is
strongly present on the top. Moreover, N at. % concentration on the
film surface is strongly reduced and in turn, it is increased to
the level in the original NbTiN film
216 212 208 204 200
5
10
15
20
25
30
35
0
Binding energy [eV]
Co
unt
s x1
03
468 464 460 456 452
30
25
20
15
10
5
0
Binding energy [eV]
Co
unt
s x1
02
-
38th International Conference of IMAPS-CPMT Poland
Rzeszw - Czarna, 21-24 September 2014
4
after surface cleaning by the Ar+ etching for 2 min. Above
results confirm presence of oxidized surface on both the as-dep. -
and annealed NbTiN film, and show differences in at.%
concentrations.
4. PHOTODETECTOR QUANTUM EFFICIENCY CHARACTERIZATION
The system quantum efficiency (SQE) of the fabricated SSPD,
based on the NbTiN film deposited on the sapphire substrate with
post annealing , was estimated. For this, the dependence of the
detector photon count rate on laser pulse intensity was measured.
The SQE was determined as a ratio of the detector count rate to the
number of photons, emitted by the laser within a linear part of
observed dependence. The laser was operated at 10 MHz repetition
rate. Linear dependence indicates that detector is in single photon
absorption regime (Fig.3a). A plateau region indicates the
background noise level which is due to so-called dark counts. The
measurement of critical temperature TC, relied on measuring of the
detector resistance as function of temperature. Fig. 3b shows the
such dependence of the normalized resistance where TC =14K was
registered. Measurements of current density on similar NbTiN film
disclosed extremely high critical current density of 12106 A/cm2
which is the best value, up to now, known for superconducting
films.
a) b)
Fig. 3. Photon count rate dependence on the laser intensity
measured on the SSPD manufactured with the NbTiN film (a);
measurement of critical temperature for the NbTiN film (NTN11shW
after RTA) (b).
5. CONCLUSIONS Quantitative analysis of atomic concentrations on
4 nm thick NbTiN film was performed by XPS investigations. The film
includes 3 at.% conc. of Ti. A deficiency of N concentration is
observed in the film because of some oxides formed on the surface.
Concentration levels of Nb-, Ti nitrides as well as contaminants
like TiOx, NbN(1-x)Ox, and Nb-oxides were evaluated. TbTiN films
deposited on Al2O3 and RTA annealed are the superior material for
SSPD, as confirmed by measurements of SSPD SQE.
REFERENCES [1] Slysz, W., Guziewicz, M., Borysiewicz, M., ,
Domagala, J.Z., Pasternak, I., Hejduk, K., Rzodkiewicz, W.,
Ratajczak, J., Bar, J., Wegrzecki, M., Grabiec, P., Grodecki,
R., Wegrzecka, I. and Sobolewski, R., Ultrathin NbN films for
Superconducting Single-Photon Detectors, Acta Physica Polonica A120
(1), 200-204 (2011).
[2] Guziewicz, M., Slysz, W., Borysiewicz, M., Kruszka R.,
Sidor, Z., Juchniewicz, M., Golaszewska, K., Domagala, J.Z.,
Rzodkiewicz, W., Ratajczak, J., Bar, J., Wgrzecki, M. and
Sobolewski R. Technology of ultrathin NbN and NbTiN films for
superconducting photodetectors Acta Physica Polonica A 120 (no.
6-A), 76 -79 (2011).
[3] Biesinger, M. C., Lau, L.W.M., Gerson, A. R., Smart, R.
St.C. Resolving chemical states in XPS analysis of first row
transition metals, axides and hydroxides: Sc, Ti, V, Cu and Zn,
Applied Surface Science (no. 257), 887-898 (2010).
[4] Darlinski, A., Halbritter, J., On the identification of
interface oxides and interface serration by ARXPS, (no. 329),
266-271 (1987).