Photon Absorption Improvement in Reststrahlen Band
WEI REN ,
and AIMIN CHANG
1.—International Joint Research Center of China for Optoelectronic and Energy Materials, School
of Physics and Astronomy, Yunnan University, Kunming 650091, China. 2.—Key Laboratory of
Functional Materials and Devices for Special Environments of CAS, Xinjiang Key Laboratory
of Electronic Information Materials and Devices, Xinjiang Technical Institute of Physics &
Chemistry, CAS, Urumqi 830011, China. 3.—Department of Physics, Shihezi University, Shihezi
832003, China. 4.—University of Chinese Academy of Sciences, Beijing 100049, China. 5.—e-mail:
email@example.com. 6.—e-mail: firstname.lastname@example.org
series ﬁlms have been fabricated on SiO
substrates by chemical solution deposition and characterized by scanning
electron microscopy, and their structural and mid-infrared (IR) properties
investigated. The results indicate slight improvement in the microstructure
and density of the ﬁlms with increasing Fe content. The results of Raman
spectroscopy showed variation in the local distortion and cation distribution at
octahedral sites with elevated Fe content. The IR optical properties of the
ﬁlms were investigated at room temperature in the wavelength range from
1.5 lmto25lm. A strong absorption peak corresponding to Reststrahlen
band located at 19.5 lm was observed and its absorption intensity found to
improve with increasing Fe content in the ﬁlms. The maximum absorption
coefﬁcient was calculated to be about 18,000 cm
. The results bear techno-
logical signiﬁcance for the design and fabrication of devices for IR detection
Key words: Thin ﬁlm, spinel, optical absorption, reststrahlen band, Raman
Over recent decades, transition-metal oxides have
been extensively investigated due to their wide
range of fascinating physical properties and various
physical phenomena, such as spin glass effects,
superconductivity, superparamagnetism, and ferro-
Among these materials, ternary
spinel oxides of type Mn-Co-Ni-O with general
are potential candidates for use in
uncooled infrared (IR) detection
due to their excep-
tional negative temperature coefﬁcient, robust ther-
mal stability, moderate resistivity, and rapid
In particular, Mn
(MCN) is considered an important composition
because its resistivity is close to the minimum
reported among Mn-Co-Ni-O ternary oxides.
over, MCN spinel is a promising material for use in
thermal detection with a wide spectral range. The
thermal, electrical, and optical properties of MCN
ﬁlm have been extensively studied, and its promis-
ing application in uncooled bolometers systemati-
However, the optical properties,
in particular the optical absorption in Reststrahlen
band, of MCN ﬁlm have not yet been completely
Dannenberg’s group reported rest-
strahlen and Raman absorption from 469 cm
from Raman scattering measurements
and Fourier-transform infrared (FTIR) spec-
as also conﬁrmed by Kong et al.
group also studied the optical properties, in partic-
ular the ultraviolet–visible–near IR and IR optical
constants, of MCN ﬁlms using spectroscopic
(Received January 4, 2017; accepted April 26, 2017;
published online May 10, 2017)
Journal of ELECTRONIC MATERIALS, Vol. 46, No. 8, 2017
Ó2017 The Minerals, Metals & Materials Society
Furthermore, iron (Fe) ion
exhibits some special features such as the change-
able valency of Fe
and ﬂexible occupation
of sublattices at A- and B-sites of the spinel
Therefore, introduction of Fe ion into
MCN spinel may result in interesting changes in
the electrical and optical properties of the material;
For example, Nikolic
´et al. reported the far-IR
optical properties of (Mn,Ni,Co,Fe)
ples using FTIR spectrometry.
In this study,
(MCNF, x= 0, 0.1, 0.15)
series spinel thin ﬁlms were synthesized on SiO
Si(100) substrates by chemical solution deposition.
The ﬁlms were characterized and their structural
and mid-IR optical properties systematically
explored. Furthermore, the effects of the Fe concen-
tration on the crystallinity, density, local distortion
and cation distribution at octahedral sites, and
intensity of absorption peak of Reststrahlen band
were systematically investigated.
MCNF spinel ﬁlms were synthesized by chemical
solution deposition. The raw solution was prepared
by mixing analytical-grade Mn(CH
(99% purity), Co(CH
O (99% purity),
O (99% purity), and FeCl
purity) in molar ratio of Mn:Co:Ni:-
Fe = 1.56:(0.96 x):0.48:x(x= 0, 0.1, 0.15; corre-
sponding samples labeled S1, S2, and S3), followed
by addition of the mixture into glacial acid solution.
The contents were then heated at 50°C and stirred
for 3 h until complete dissolution of the mixture.
The concentration of the ﬁnal solution was adjusted
to 0.15 M. The precursor solution was then ﬁltered
through a 0.45-lm syringe to remove residual
particulates from the solution. The solution was
spin-coated on SiO
/Si(100) substrates at 500 rpm
for 5 s and 4000 rpm for 20 s, followed by drying in
a furnace at 350°C for 5 min. The spin-coating and
drying procedures were repeated for ten times, then
the samples were annealed in air at 750°C for 1 h,
which resulted in formation of MCNF spinel oxide
The crystalline phase of the thin-ﬁlm materials
was identiﬁed by x-ray diffraction (XRD, Bruker
D8). The ﬁlm morphology was characterized by
scanning electron microscopy (SEM, Supra55VP,
Zeiss). To verify the distribution of various elements
in the ﬁlms, energy-dispersive x-ray (EDX) spectra
and elemental maps were obtained using an EDX
spectroscope (LEO 1430VP). Raman spectra were
measured using a Raman spectrometer (NRS-1000)
laser (20 mW, 532 nm) as excitation source
in backscattering conﬁguration. The optical proper-
ties were assessed by SE (SENTECH SE850) in the
IR spectral range at angle of incidence of 70°and
RESULTS AND DISCUSSION
The XRD patterns of the MCNF series ﬁlms are
shown in Fig. 1, where the peaks were labeled by
referring to the standard Joint Committee on
Powder Diffraction Standards (JCPDS) card for
(No. 01.1110). The MCNF ﬁlms exhibited
polycrystalline structure with preferred (311) ori-
entation. The XRD patterns clearly indicate that
substitution of Co with Fe did not lead to a change
in the original spinel structure. This can be
attributed to the fact that the ionic radius of Fe
(0.0645 nm) is close to that of Co
(0.061 nm) and
at octahedral site.
The peaks of
the Fe-containing ﬁlms showed slight blue-shift
compared with those of pure MCN ﬁlm, indicating
increase of the lattice constants. This lattice expan-
sion can be explained as follows: for MCN spinel, a
typical cation distribution might be
at octahedral site, as
well as more Fe
(0.078 nm) replacing Co
(0.0745 nm) at the tetrahedral site, the calculated
lattice constants increase from 8.308 A
˚for S1, to
˚for S2, to 8.331 A
˚for S3. Moreover, notably,
the peak widths also broadened from S1 to S3,
indicating shrinkage of the MCNF grains. The
average crystallite sizes calculated by using the
Scherrer equation for the (311) peak tended to
decrease from 32.970 nm for S1, to 28.665 nm for
S2, to 25.086 nm for S3, indicating degradation of
Figure 2shows the surface morphology of the
MCNF series ﬁlms, revealing the existence of
particle chains and pores for S1.
particle chain/pore structure tended to disappear on
addition of Fe. The average particle size decreased
from S1 to S3. The ﬁlm density and surface mor-
phology correspondingly improved, which could
help minimize the effects of ambient humidity or
Fig. 1. XRD patterns of MCNF series ﬁlms.
Zhang, Shi, Ren, Zhou, Lu, Bao, Wang, Bian, Xu, and Chang5350
absorption of oxygen from the surrounding air onto
the ﬁlm. The insets in Fig. 2show cross-sectional
SEM images of the thin ﬁlms, with corresponding
ﬁlm thickness values of around 400 nm, 395 nm,
and 390 nm for S1, S2, and S3, respectively. The
difference in the ﬁlm thickness values is small and
will not affect the optical properties of the ﬁlms as
measured by SE.
EDX spectroscopy was performed to quantify the
composition of the samples. The atomic concentra-
tions of Mn, Co, Ni, and Fe in the series of ﬁlms
were obtained and are presented in Table I. Based
on these values, the Mn:Co:Ni(:Fe) ratio was calcu-
lated to be 50:33:17 for S1, 50:32:16(:2) for S2, and
50:29:17(:4) for S3.
Figure 3clearly demonstrates that the surface
morphology of ﬁlm S3 was the least defective.
Elemental distribution maps of Mn, Co, Ni, and Fe
for ﬁlm S3 are shown in Fig. 3. These maps conﬁrm
that the distribution of these elements was roughly
uniform, except for some holes randomly distributed
on the surface of ﬁlm S3.
Figure 4shows the Raman spectra for all the
MCNF ﬁlms at room temperature. According to
the strongest peak at
is assigned to A
vibration mode, which
originates from the symmetric B–O stretching
vibration and reﬂects the local lattice bonding of
. The peak centered at 490 cm
assigned to the F
vibration modes, which reﬂects
the symmetric bending vibration of Mn
–O bond at
octahedral site. The peak at 820 cm
is ascribed to
Si–O bond, which comes from the substrate. Clearly,
the intensity of A
peak improves with Fe addition
and its position shifts slightly toward higher
wavenumber from S2 to S3. When Fe
place of Co
, an increase in the length of the B–O
bond brings about a reduction of the bond strength
and the corresponding red-shift of the Raman
The radius of Fe
(0.0645 nm) is
similar to that of Mn
(0.0645 nm), compared with
that of Co
(0.061 nm), thus inclusion of more Fe
at octahedral site results in reduction of the John–
Teller lattice distortion, as conﬁrmed by the
increased peak intensity and narrower peak width
mode. The intensity and position of F
remained stable at elevated Fe content, indicating
insigniﬁcant effect of Fe
addition on the content of
SE was used to obtain the optical properties of the
MCNF series ﬁlms. SE measures the amplitude u
and phase D, which are related to the optical
constants by the equation q¼tanðuÞeiD, where qis
the complex reﬂectivity ratio. From the ellipsomet-
ric measurements, the pseudodielectric function
(e¼e1þie2) can be obtained by using the following
where /is the angle of incidence.
Fig. 2. SEM images and cross-sectional SEM images (inset) of MC
series ﬁlms with (a) x= 0, (b) x= 0.1, and (c) x= 0.15.
Photon Absorption Improvement in Reststrahlen Band of Mn
To attain precise optical constant curves for the
thin ﬁlms from ellipsometry, the data processing is
critical. An appropriate model helps signiﬁcantly in
ﬁtting the data. In this work, a Drude–Lorentz
model of the MCNF ﬁlm was selected because it can
precisely describe the contribution of both free and
bound electrons responsible for the optical proper-
ties. A three-layer (air/MCNF/SiO
) structure was
used to describe the samples, and the mid-IR optical
properties of the MCNF ﬁlms are shown in Fig. 6.
The Drude–Lorentz oscillator formula below
approximately describes the dispersion relation of
the complex dielectric function (e):
, and e
represent the high-frequency
dielectric constant, and the real and imaginary
parts of the dielectric constant, respectively.
=kreﬂects the inﬂuence of the internal
electric ﬁeld on the Lorentz oscillator.
pand xs¼e=umshow the relation-
ship among the concentration Nof the free carriers,
the mobility lof the free carriers, and the effective
of the free carriers. In the last part, X
represent the amplitude, center frequency,
and damping factor of the Lorentz oscillator, respec-
tively. The dielectric constant can then be trans-
formed to optical properties (refractive index nand
extinction coefﬁcient k) by using the following
Table I. EDX spectroscopy results for MCNF series thin ﬁlms
S1 S2 S3
at.% Error (%) at.% Error (%) at.% Error (%)
Mn 50.45 1.65 49.74 1.27 49.95 1.19
Co 32.98 1.34 31.58 1.05 28.83 0.99
Ni 16.57 0.85 16.27 0.61 16.67 0.69
Fe – – 2.41 0.22 4.55 0.25
Fig. 3. EDX spectroscopy surface elemental maps for ﬁlm S3.
Zhang, Shi, Ren, Zhou, Lu, Bao, Wang, Bian, Xu, and Chang5352
The mean square error (MSE) is a measure of the
ﬁtting result, described as follows:
where Land Mare the ﬁtting points and measured
modeled and measured ellipsometric angles, respec-
tively; and r
are the standard deviations of
the measured ellipsometric angles. Figure 5shows
the measured points and ﬁt lines overlapping each
other for all the ﬁlms in the series. Furthermore, the
detailed ﬁtting parameters of the Drude–Lorentz
oscillator formula are listed in Table II.
The refractive index nand extinction index kwere
determined from uand D. Figure 6a shows that n
generally dropped with increasing wavelength from
1.5 lmto16lm for all the ﬁlms, dramatically
increased from 16 lmto21lm due to anomalous
dispersion, then remained in the range of 3.4 to 3.7
at wavelength of 21 lmto25lm. Notably, the
decreasing tendency of nfor wavelength of 1.5 lmto
16 lm is similar to that reported in literature,
Fig. 4. Raman spectra of MCNF series thin ﬁlms.
Fig. 5. Experimental and ﬁt SE spectra (/and D)ofMC
series ﬁlms with (a) x= 0, (b) x= 0.1, and (c) x= 0.15.
Photon Absorption Improvement in Reststrahlen Band of Mn
although that measurement stopped at 7 lm. More-
over, the nvalues reported in literature are higher
than those of S1, which may be due to the different
ﬁlm deposition method. On addition of Fe to the
ﬁlms, the nvalues of S2 and S3 became comparable
to that reported in literature.
The improvement in
nvalue with increasing Fe content for S2 and S3
can be ascribed to the signiﬁcant reduction of pores
in these ﬁlms, which correspondingly reduces the
effective dielectric constant.
Another reason may
be the lower grain size of the Fe-containing ﬁlms.
The minimal value at 16 lm is due to the existence
of a powerful dielectric loss function.
Figure 6b shows the kcurves of the MCNF ﬁlms
in the wavelength range from 1.5 lmto25lm. A
signiﬁcant absorption region is centered at 19.5 lm
and covers a broad region of 16 lmto22lm, which
can be attributed to reststrahlen absorption.
et al. reported four IR-active frequencies of MCN
(one of them, 555 cm
, corresponds to the
wavelength of 18.1 lm).
´et al. also reported
the occurrence of two vibration modes (at around
and 470 cm
, corresponding to wave-
lengths of 16.1 lm and 21.3 lm) for MCN spinel
The intensity of the highest absorption
apparently increased from S1 to S3, probably due
to the following reasons: ﬁrst, the absorption peak
corresponding to Fe
–O bonds is also found
between 470 cm
and 620 cm
The content of
Fe in the ﬁlms generally increased, while that of Mn
remained constant. Obviously, the optical absorp-
tion between 470 cm
and 620 cm
with addition of more Fe to the ﬁlm. Second, the
variation of the cation distribution coupled with the
reduced lattice distortion caused by the replacement
, as evidenced by the Raman
results, may result in improved reststrahlen absorp-
tion. Lutz et al. pointed out that all the allowed IR
modes represent typical lattice vibrations resulting
from the contribution of all atoms and forces of the
Table II. Fitting parameters of Drude–Lorentz oscillator formula
S1 S2 S3
Thickness (nm) 200 157.26 191.62
e10.96 0.99 0.84
10.001 0.016 0.071
xp(1/cm) 1004.2 892.98 1017.54
xs(1/cm) 351.53 0.35 36.59
Xo(1) (1/cm) 12,523.33 13,599.79 12,919.55
Xp(1) (1/cm) 4591.65 3695.59 4600.17
Xs(1) (1/cm) 4500.07 420.02 8.09
Xo(2) (1/cm) 452.24 5904.88 1725.74
Xp(2) (1/cm) 816.94 271.93 1715.38
Xs(2) (1/cm) 57.23 156,544.30 2056.74
Xo(3) (1/cm) 1130.47 1599.63 3282.88
Xp(3) (1/cm) 1276.36 1506.57 2070.23
Xs(3) (1/cm) 1004.25 2676.64 1929.69
MSE 1.37 1.86 1.73
Fig. 6. (a) Refractive index nand (b) extinction coefﬁcient kcurves of MCNF series ﬁlms in mid-infrared band.
Zhang, Shi, Ren, Zhou, Lu, Bao, Wang, Bian, Xu, and Chang5354
spinel structure, though some of them can be more
affected by the metal ions at octahedral or tetrahe-
However, further studies are deﬁnitely
needed for comprehensive understanding of the
distribution of cations due to Fe doping. Finally,
the improved ﬁlm quality proved by SEM images
can promote the absorption intensity. Addition of Fe
to the MCN ﬁlms generated smaller grains, facili-
tated densiﬁcation of the thin ﬁlm, and improved
the ﬁlm morphology, thus increasing oscillators at
the absolute volume. Participation of more oscilla-
tors in lattice vibrations leads to occurrence of more
prominent optical absorption. Moreover, the absorp-
tion coefﬁcient can be calculated by using the
formula a=4pk/k, where kis the wavelength. Based
on this formula, the strongest absorption coefﬁcient
is 17,974 cm
, calculated for S3 at 19.5 lm.
MCNF series ﬁlms were synthesized on SiO
Si(100) substrate by chemical solution deposition.
Pure MCN ﬁlm was polycrystalline structured with
poriferous surface; however, increasing Fe content
coupled with decrease of Co in the ﬁlms helped
smoothen the ﬁlm morphology and improved the
reststrahlen absorption, although the crystallinity
degraded. The strong absorption peak located at
19.5 lm is attributed to characteristic lattice vibra-
tions. The heightening of the peak with Fe doping
was due to the increasing number of oscillators or
the disorder of the lattice structure. The maximum
value of awas approximately 18,000 cm
of Fe to MCN ﬁlm modiﬁed the microstructure, ﬁlm
densiﬁcation, lattice distortion, and cationic distri-
bution, and helped to improve the reststrahlen
This work was supported by the Thousand Youth
Talents Plan (Y42H831301), National Natural Sci-
ence Foundation of China (61504168), West Light
Foundation of the Chinese Academy of Sciences
(XBBS-2014-04), Science and Technology Program
of Hebei Province (D2016403064, 16044601Z), He-
bei Outstanding Young Scholars and Hebei Science
and Technology Support Program (15211121), and
Foundation of Director of Xinjiang Technical Insti-
tute of Physics & Chemistry, CAS (2015RC010).
1. W. Zhou, J. Wu, C. Ouyang, Y.Q. Gao, X.F. Xu, and Z.M.
Huang, J. Appl. Phys. 115, 093512 (2014).
2. C. Ouyang, W. Zhou, J. Wu, Y. Hou, Y.Q. Gao, and Z.M.
Hang, Appl. Phys. Lett. 105, 022105 (2014).
3. C. Ma, H.G. Wang, P.J. Zhao, J.B. Xu, A.M. Chang, and L.
Wang, Mater. Lett. 136, 225 (2014).
4. Y. Hou, Z.M. Huang, Y.Q. Gao, Y.J. Ge, J. Wu, and J.H.
Chu, Appl. Phys. Lett. 92, 202115 (2008).
5. C. Ouyang, W. Zhou, J. Wu, Y.Q. Gao, F. Zhang, and Z.M.
Huang, Mater. Res. Innov. 19, S7 (2015).
6. W.W. Kong, L. Chen, B. Gao, B. Zhang, P.J. Zhao, G. Ji,
A.M. Chang, and C.P. Jiang, Ceram. Int. 40, 8405 (2014).
7. R. Dannenberg, S. Baliga, R.J. Gambino, A.H. King, and
A.P. Doctor, J. Appl. Phys. 86, 2590 (1999).
8. M.V. Nikolic
´, K.M. Paraskevopoulos, O.S. Aleksic
Zorba, S.M. Savic
´, V.D. Blagojevic
´, D.T. Lukovic
´, and P.M.
´,Mater. Res. Bull. 42, 1492 (2007).
9. R.N. Jadhav, S.N. Mathad, and V. Puri, Ceram. Int. 38,
10. S.M. Savic
´a, M.V. Nikolic
´, K.M. Paraskevopoulos, T.T.
Zorba, N. Nikolic
´, V. Blagojevic
´, O.S. Aleksic
´, and G.
´,Ceram. Int. 39, 1241 (2013).
11. Y.Q. Gao, Z.M. Huang, Y. Hou, J. Wu, Y.J. Ge, and J.H.
Chu, Appl. Phys. Lett. 94, 011106 (2009).
12. C. Ma, W. Ren, L. Wang, L. Bian, J.B. Xu, and A.M. Chang,
Mater. Lett. 153, 162 (2015).
13. C. Ma, W. Ren, L. Wang, J.B. Xu, A.M. Chang, and L. Bian,
J. Eur. Ceram. Soc. 36, 4059 (2016).
14. J.Y. Wang and J.J. Zhang, J. Mater. Res. 27, 928 (2012).
15. L. Malavasi, P. Galinetto, M.C. Mozzati, C.B. Azzoni, and
G. Flor, Phys. Chem. Chem. Phys. 4, 3876 (2002).
16. M. Grimsditch, Phys. Rev. Lett. 52, 2379 (1984).
17. S. D’Elia, N. Scaramuzza, F. Ciuchi, C. Versace, G. Strangi,
and R. Bartolino, Appl. Surf. Sci. 255, 7203 (2009).
18. R. Dannenberg, S. Baliga, R.J. Gambino, A.H. King, and
A.P. Doctor, J. Appl. Phys. 86, 514 (1999).
19. H.D. Lutz, B. Mu¨ ller, and H.J. Steiner, J. Solid State
Chem. 90, 54 (1991).
Photon Absorption Improvement in Reststrahlen Band of Mn