Optica Applicata, Vol. L, No.

1, 2020
DOI: 10.37190/oa200106

Utilization of the cyclic interferometer

in polarization phase-shifting technique
to determine the thickness of transparent thin-films
RAPEEPAN KAEWON1, CHUTCHAI PAWONG2, RATCHAPAK CHITAREE3,
TOSSAPORN LERTVANITHPHOL4, APICHAI BHATRANAND5*

1
Silpakorn University, Faculty of Engineering and Industrial Technology,
Department of Electrical Engineering, 6 Rajamankha Nai Rd, Nakornpathom 73000, Thailand
2
Rajamangala University of Technology Krungthep, Faculty of Science and Technology,
Physics Division, 2 Nanlinji Road, Bangkok 10120, Thailand
3
Mahidol University, Faculty of Science, Department of Physics,
272 Rama VI Road, Bangkok 10400, Thailand
4
National Electronics and Computer Technology Center, Optical Thin-Film Laboratory,
Pathumthani 12120, Thailand
5
King Mongkut’s University of Technology Thonburi, Faculty of Engineering,
Department of Electronic and Telecommunication Engineering,
126 Pracha U-thid Road, Bangkok 10140, Thailand
*Corresponding author: apichai.bha@mail.kmutt.ac.th

An alternative polarization phase-shifting technique is proposed to determine the thickness of

transparent thin-films. In this study, the cyclic interferometric configuration is chosen to maintain
the stability of the operation against external vibrations. The incident light is simply split by
a non-polarizing beam splitter cube to generate test and reference beams, which are subsequently
polarized by a polarizing beam splitter. Both linearly polarized beams are orthogonal and counter
-propagating within the interferometer. A wave plate is inserted into the common paths to introduce
an intrinsic phase difference between the orthogonal polarized beams. A transparent thin-film sam-
ple, placed in one of the beam tracks, modifies the output signal in terms of the phase retardation
in comparison with the reference beam. The proposed phase-shifting technique uses a moving mir-
ror with a set of “fixed” polarizing elements, namely, a quarter-wave retarder and a polarizer, to
facilitate phase extraction without rotating any polarizing devices. The measured thicknesses are
compared with the measurements of the same films acquired using standard equipment such as the
field-emission scanning electron microscope and spectroscopic ellipsometer. Experimental results
with the corresponding measured values are in good agreement with commercial measurements.
The system can be reliably utilized for non-destructive thickness measurements of transparent thin
-films.

Keywords: polarization phase-shifting technique, cyclic interferometric configuration, non-destructive

thickness measurements, transparent thin-films.
70 R. KAEWON et al.

1. Introduction
Phase-shifting interferometry is generally composed of two major procedures – the
phase shifting step and the phase extraction algorithm – in which the latter gives the
phase information of a sample introduced in the interferometric system. The most com-
mon methods to generate phase shift are moving a mirror [1– 4], displacing a diffraction
grating [5], tilting a glass plate [6, 7], and rotating polarizing devices such as a phase
plate [8–12], and a polarizer [13–15]. The choice of the phase-shifting method for
a particular application depends on available optical components coupled with its ad-
vantages deemed appropriate for the implementation. In this study, a cyclic interfero-
metric configuration, which is less prone to the instability due to environmental
disturbances, is chosen [16, 17]. The proposed phase-shifting technique used to gen-
erate phase shifts for the setup is based on the mirror displacement and the combination
of a quarter-wave retarder and a polarizer. By integrating both units into the cyclic in-
terferometer, an interesting feature, which cannot be obtained from either mirror dis-
placement or the combination of a quarter-wave retarder and a polarizer, is achieved.
Two orthogonal linearly polarized beams emitted from the cyclic interferometer are
transformed by the combination of a quarter-wave retarder and a polarizer fixed at
a pre-aligned orientation into two circularly polarized beams with opposite handed-
ness. They then combine to give a linearly polarized beam whose electric vector has
a particular orientation relative to a reference axis. It is well known that the modulated
phase difference between the two circularly polarized beams leads to the rotation of
the resultant polarized beam: the so-called rotating polarized light. The modulation can
be achieved with the moving mirror attached to the piezoelectric transducer (PZT)
whose speed can simply be controlled by the driving signal. The proposed technique
results in the rotation of the signal outputs via modulation while the polarizing device,
either a wave-phase retarder or a polarizer, is fixed. This actually is contrary to the normal
procedure of the polarization phase-shifting method whereby either the wave-plate
retarder or polarizer has to be rotated to achieve the phase shift.
To implement the proposed polarization phase-shifting technique, a homogeneous
and transparent thin-film of WO3 is used as a test sample. The sample inserted in the
optical path between the two mirrors, i.e. M1 and PZT(M2), directly introduces a phase
shift in the reference beam. With the proposed extraction technique, which will be dis-
cussed later, the phase shift can simply be determined and subsequently converted into
the corresponding thickness of the thin-film sample. To verify the obtained thickness,
the sample thickness was also examined with thin-film measuring instruments such as
a field-emission scanning electron microscope and a spectroscopic ellipsometer.

2. Principle
2.1. Principle of a cyclic interferometer system without a sample
The schematic of the cyclic interferometer (CI) system is illustrated in Fig. 1. The laser
light source (LS) emits linearly polarized light, whose direction is parallel to the axis
Utilization of the cyclic interferometer in polarization phase-shifting technique... 71

M1

WP

S1
Path 1
S2

PB
D1
D2
PZT(M2)
Path 2 Beam 1 Beam 2
QWP P
HWP
BSC

Parallel component
Perpendicular component LS

Fig. 1. Schematic of a cyclic interferometer.

of the polarization beam splitter (PB), reaching the CI. A beam splitter cube (BSC)
placed after the laser is used to split an incident light beam into beam 1 and beam 2.
Both beams can be rotated to any commonly desired orientation by adjusting the half-
wave plate (HWP). The HWP is oriented at 22.5° with respect to the axis of the PB.
This results in the polarization directions of both beams being at 45° with respect to the
axis of the PB. Therefore, the Jones vectors, representing the electric fields of beam 1
and beam 2, are identical and given as

1 1
E i = ------------ (1)
2 1
At the PB, both polarized beams are equally split into two sets of orthogonal linearly
polarized beams. Each polarized beam travels in perpendicular directions as guided by
mirror 1 (M1) and mirror 2 (M2) that is attached to a PZT (PZT(M2)). The states of the
twin polarized light beams with an electric field, as the parallel transmitted light beams
travelling along path 1, are identically linearly parallel to the plane of the interferometer.
In contrast, the reflected light beams with an electric field propagating along path 2, are
identically linearly perpendicular to the plane of the interferometer. These dual polarized
light beams from the two paths are then recombined at the PB and are transformed into
two sets of circularly polarized light beams by the quarter-wave plate (QWP). The re-
sulting twin beams are the so-called dual rotating polarized light (RPL). Both output
beams, with electric fields Ei T and Ei R, in the two tracks can be equivalently expressed as
72 R. KAEWON et al.

1 exp  iδ F 
E iT = QW T PBT M R WP M R T PBT E i = ------ (2)
2 – i exp  iδ 
F

– i exp i  δ M + δ M  t  + δ S 
1 1 2
E iR = QW T PBR M R WP M R T PBR E i = ------ (3)
2
exp i  δ M + δ M  t  + δ S 
1 2

Here, TPBT and TPBR denote the Jones matrices of the linear PB [18, 19]. The phase
shifts δ of the dual polarized light beams after reflections from mirror M1 and PZT(M2)
are δM1 and δM2(t), where δM2(t) is time-dependent. The Jones representation MR of each
mirror, a linear birefringent wave plate (WP), and the output quarter-wave plate (QWP)
QW are given by following equations:

1 0
MR = (4)
0 exp  iδ 

exp  iδ F  0
WP = (5)
0 exp  iδ S 

1 1 –i
QW = ------------ (6)
2 –i 1
where δF and δS are phase retardations represented by the fast and slow eigenmodes [20],
respectively, and the QWP is oriented with its fast axis at – 45° with respect to the trans-
mission axis of the PB. Note that the vibrating PZT(M2) is sinusoidally driven by a func-
tion generator, and the dual phase shifts introduced by M2 alternate with time. Generally,
each pair of twin polarized light beams is equally vibrated by a PZT, and they are in
phase. Hence, a general WP is placed between M1 and PZT(M2) to produce an unbal-
anced phase shift through its fast and slow axes, so that the relative phases are changed
within the CI system. Both the output beams have the same electric field Eout that is
equivalent to the summation of Ei T and EiR in the two tracks in the absence of a sample,
and can be written as [19]

1 1 1 π 
E out = ------  exp  iδ F  + exp i  δ S + δ M  t  + δ M – ------ 
2  –i i  2 1 2 

sin  δ t  t   i π
= exp ------  δ F + δ S + δ M  t  + δ M – ------ (7)
cos  δ t  t   2  2 1 2

with δt(t) = δS +δM1 + δM2(t) – δF representing the dual output phase shifts in the CI.
Here, it is obvious that the mirror movement gives rise to the change in the phase shifts
Utilization of the cyclic interferometer in polarization phase-shifting technique... 73

corresponding to the varying orientations of the dual RPL. Subsequently, the two output
light beams after the QWP reach the polarizer (P), and are separately detected by photo-
detector 1 (D1) and photodetector 2 (D2). The dual RPLs are automatically dual rotating
with the fixed polarizer; the output intensities detected by D1 and D2 are expected to
follow the variation described by Malus’ law. Here, we note that the orientations of
the parallel-polarized light beams are “rotating” with respect to the transmission axis
of the polarizer. Hence, any phase shifts of interest will be introduced later by inserting
WO3 thin-film samples on a BK7 glass substrate and purely BK7 glass substrate in the
cavity composed of a sample track and reference track, respectively.

2.2. Principle of CI with the presence of a sample

The dual beams within the CI are employed as a set of the probing tools to investigate
the phase shifts introduced by a sample placed between M1 and PZT(M2), as depicted
in Fig. 1. The substrate of the sample is BK7. The WO3 thin film is deposited on half
of the BK7 substrate, and this sample part is called S1. The other half S2 is just a pure
BK7 substrate. The sample is positioned in the CI system such that tracks associated
with S1 and S2 are called sample and reference tracks, respectively. In the sample track,
the electric fields of the polarized beams traveling along paths 1 and 2 are denoted by
Ei T, S1 and Ei R, S1, respectively. The Jones matrix S of the homogenous sample is ex-
pressed as [18, 19]

T1 0
S = (8)
0 T 2 exp  iΔ S i 

where T1 and T2 represent the transmission coefficients along the principal axes of the
thin-film on the transparent sample and ΔSi is the phase retardation introduced by the
Si track when S1 and S2 are inserted, respectively. Thus, the Jones vector Ei T, S1 repre-
senting the transmitted light in the S1 track can be expressed as

1 T1
E iT S = ------- exp  iδ F  (9)
1 2 – iT
1

Likewise, EiR, S1, the Jones vector of the reflected light beam in the S1 track, is ex-
pressed as

1 – iT 2
E iR S = ------- exp i  Δ S + δ S + δ M  t  + δ M  (10)
1 2 T 1 2 1
2
θ
Therefore, the Jones vector E out  S 1 of the S1 output beams, emerging from the CI setup,
can be expressed as
θ
E out  S = P  θ  E iT S + E iR S
1 1 1
(11)
74 R. KAEWON et al.

where P(θ) represents the Jones matrix of the linear polarizer P whose transmission
axis is aligned at an angle θ with respect to the reference axis [20, 21]. In this study,
the polarizer is adjusted to align at an angle of 45° to the reference axis. The RPL ori-
entation at 0°, 45°, and 90° can be expressed as

0° T 12 T 1 T 2 sin  Δ S + δ t  t   T 22
1
I out S1 = ---------
- + -------------------------------------------------------
- + ---------
- (12)
4 2 4

45° T 12 T 1 T 2 cos  Δ S + δ t  t   T 22
1
I out S1 = ---------- + --------------------------------------------------------- + ---------- (13)
4 2 4

90° T 12 T 1 T 2 sin  Δ S + δ t  t   T 22
1
I out S1 = ---------
- – -------------------------------------------------------
- + ---------
- (14)
4 2 4
Thus, the phase retardation ΔS1, introduced by an inserted WO3 thin-film sample
in the S1 track, can be obtained from Eqs. (12) to (14) as

 0°
I out 90°
 S 1 – I out S 1

–1 
-
Δ S + δ t  t  = tan --------------------------------------------------------------- (15)
1  2I 45° – I 90° – I 0° 
 out S1 out S 1 out S 1 
The RPL orientation depends on the mirror vibration. Subsequently, the output light
after QWP goes into the polarizer P and the photodetector D1. The rotation of the linearly
polarized output beam with respect to a static linear polarizer automatically allows the
orientation of the beam caused by the initial phase from the apparatus to be offset. This
action is marked by a time called the offset time, which indicates a new starting time
for a subsequently introduced phase shift. Therefore, the both beams of phase shifts
will be introduced later by a thin-film sample which is measured relative to the offset
times. The initial phase shift δt(t) can then be offset. The polarization state after QWP
is rotated from its original orientation as a result of the additional retardation caused
by the sample. Thus, Eq. (15) can be rewritten as

 0°
I out 90°
 S 1 – I out S 1

–1 
ΔS -
= tan --------------------------------------------------------------- (16)
1  2I 45° – I 90° – I 0° 
 out S 1 out S1 out S 1 
By carrying out the derivation in a similar manner, the phase retardation caused by
the presence of the pure BK7 substrate in the S2 track can be expressed as

 0°
I out 90°
 S 2 – I out S 2

–1 
ΔS -
= tan --------------------------------------------------------------- (17)
2  2I 45° – I 90° – I 0° 
 out S 2 out S 2 out S 2 
Finally, we can find the phase difference between those two tracks by subtracting
Eq. (17) from Eq. (16) as
Utilization of the cyclic interferometer in polarization phase-shifting technique... 75

 0°
I out 90°
 S 1 – I out S 1
  0°
I out 90°
 S 2 – I out S 2

–1   –1  
ΔS = tan ---------------------------------------------------------------
- – tan ---------------------------------------------------------------
- (18)
 2I 45° – I 90° – I 0°   2I 45° – I 90° – I 0° 
 out S 1 out S 1 out S  1  out S 2 out S 2 out S  2

Note that the simultaneous signal from the reference track (the beam passing a pure
BK7 substrate) provides both initial orientation and phase, which are essentially used
to identify intensities at specific orientations and work out the phase shift introduced
by the thin-film sample.

2.3. Film thickness determination

The transmitted light T under various conditions can be carefully examined [22]. The phase
retardation ΔS, as a function of the film thickness, can be given by 4 πn1 d1 /λ, where
n1 is the refractive index of the deposited film (2.10 for WO3) [23], d1 represents the
film thickness, and λ denotes the light wavelength (632.8 nm). Because the phase re-
tardation ΔS obtained from Eq. (18) is in the form of a tangent function that repeats
periodically every 180° or π radians, the phase shift satisfying the condition can be
written in terms of the order number m as Δm = ΔS + m π. Further, the phase shift ΔS in
Eq. (18) can be accumulated or increased to match the measured ΔS by adding integer
multiples of π, represented as mπ. Here, m denotes the cumulative integer chosen as the
sign changes from negative to positive, i.e., it specifies the number of periods of the
tangent function of ΔS. Prior to calculating the thickness d1, this relation must be satis-
fied to determine the exact order of the periodic tangent function of ΔS [24]. The pa-
rameter dcutoff is employed to determine the appropriate value of the order number m
when the deposition time td and approximate deposition rate Rd are known. In this
study, Rd from the commercial UHV sputtering system used (AJA International Inc.,
ATC 2000-F) is 30 nm/min. The cutoff thickness dcutoff , at the point at which the second
cycle begins (ΔS = π), can be expressed as λ /4n1. Accordingly, the appropriate value
of m can be determined using td Rd/dcutoff. For WO3 with a refractive index of 2.10 and
a red laser having wavelength of 632.8 nm, dcutoff is found to be 70 nm. According to
the information, the values of m at 10, 15, and 20 min with rate Rd = 30 nm/min are 4,
6, and 8, respectively. Eventually, the thin-film thickness d1 can be determined with
an appropriate value of m as [19]

Δm λ
d 1 = -----------------
- (19)
4π n 1

3. Signal treatment algorithm to evaluate phase retardation

Signal processing was employed to determine both the phase shifts given by Eqs. (16)
and (17) between the parallel-polarized light beams in the two tracks. A digital oscil-
loscope (Yokogawa DLM2000) was used to obtain the desired information. The opti-
mized amplitude and frequency of a modulating sinusoidal trigger signal were set at
76 R. KAEWON et al.

1.6 V and 41 Hz, respectively. The trigger function from the oscilloscope was carefully
adjusted to stabilize signals from reference and sample tracks to be viewed clearly on
the screen. We then produced the unbalanced phase by inserting a general WP inside the
twin paths. Figure 2 shows both the signal intensities acquired when WO3 thin-film
samples on BK7 substrate and bare BK7 substrate are inserted. At the output of the CI,
the amplitude and phase information from the two tracks are related to the polarizer,
which is oriented at 45° with respect to the PB axis. Upon inserting thin-film samples

I45 ° Reference track S2 a

10 out, S2 Sample track S1
8 °
I0out,
8 S1

6 6
° 45°
I90
CH1 [ V ]

out, S2 I out, S1
4 4 °
I0out, S2

2
2
0
0 I90 °
out, S1
–2

I45 ° Reference track S2 b

10 out, S2 Sample track S1
8
8
°
I0out, S1
6 6
I45 °
CH1 [ V]

I90 ° out, S1
out, S2

4 4 °
I0out, S2

2 2

0 I90 °
out, S1
0
–2

I45 ° Reference track S2 c

10 out, S2 Sample track S1

8
8
°
I0out, S1
6 6
CH1 [V ]

I90 °
out, S2
°
I0out,
4 4
S2

I45 °
2 out, S1
2
0
0 I90 °
out, S1
–2
–10 –5 0 5 10 –2.5 0.0 2.5
Time [ms] Time [ms]

Fig. 2. Comparative differential relative phases of the output intensities of reference and sample waveforms
for various deposition times: 10 min (a), 15 min (b), and 20 min (c).
Utilization of the cyclic interferometer in polarization phase-shifting technique... 77

with deposition times of 10, 15, and 20 min into two tracks between M1 and PZT(M2),
the output intensities of the parallel-polarized light are directly perturbed in terms of
both amplitude and phase. This perturbation gives rise to the differential amplitudes
and phase intensities between the two output signals. The sample and reference output
waveforms are analyzed and the positions corresponding to three intensity points are
checked. The orientation of the dual RPL is automatically rotated by θ due to the PZT
modulation. The specific intensities of the two tracks can then be clearly observed when
the polarizer is set at 45° with respect to the PB axis. The outputs of the light intensities
0° 45° 90°
of the two sets are denoted by sample waveforms I out  S 1 , I out S 1 , and I out S 1 , and ref-
0° 45° 90°
erence waveforms I out  S 2 , I out S 2 , and I out S 2 . Figures 2a–2c depict waveforms ob-
tained after inserting a thin-film sample. Equation (18) is employed to evaluate the
differential phase retardations ΔS that can be converted into the thin-film thicknesses
using Eq. (19).

4. Experimental results and discussion

The degree of polarization (DOP) values of the dual RPL from the two tracks of the CI
were determined to ensure the quality of the generated rotating polarized beams, which
is very crucial for thin-film thickness measurements. A signal algorithm program was
utilized to obtain the differential phase retardation (later converted to the film thick-
nesses). The film thickness read-outs obtained from the CI setup were compared with
the ones acquired using conventional techniques such as variable-angle spectroscopic
ellipsometry (VASE; J.A. Woollam) and field-emission scanning electron microscopy
(FE-SEM; Hitachi S-4700) measurements.

4.1. Film thickness determination

The expected twin output light beams from both tracks of the CI are identically linearly
polarized beams. Ideally, the perfectly polarized light beams obtained from both outputs
have a degree of polarization (DOP) of 100%. The characteristics of the dual beams were
evaluated in terms of the four Stokes parameters. The output beam intensities yielded
four terms of the Stokes parameters for the sample track as S0 = 11.237 ± 0.080 V,
S1 = 11.210 ± 0.080 V, S2 = –0.028 ± 0.114 V, and S3 = 0.009 ± 0.035 V. The correspond-
ing values for the reference track were S0 = 11.134 ± 0.066 V, S1 = 11.116 ± 0.064 V,
S2 = – 0.039 ± 0.069 V, and S3 = 0.015 ± 0.015 V. These parameters were processed to
obtain the DOP [18] of D1 and D2 for the generated dual RPL as 99.72% and 99.83%
for sample and reference tracks, respectively. These results clearly indicate that the
states of the output dual beams from the CI system are extremely close to those of the
ideal linearly polarized light beams. Subsequently, WO3 thin-film samples of different
thicknesses were characterized using the CI setup. The samples of interest were insert-
ed in the light-beam path of the sample track between PZT(M2) and M1. Both the output
light intensities detected by D1 and D2 in terms of the voltages were processed in the
algorithm described in Section 3, so that the phase shifts ΔS1 and ΔS2 needed could be
achieved. The three output light intensities listed in the Table are the average values
78 R. KAEWON et al.

T a b l e. Average maximum intensities under various study conditions.

Average maximum intensity [V]

Sample 0° 0° 45° 45° 90° 90°
I out S1 I out  S2 I out  S1 I out  S2 I out  S1 I out  S2
WO3 = 10 min 7.63 4.53 6.18 8.75 0.52 4.40
WO3 = 15 min 6.38 4.75 6.25 8.84 0.94 4.62
WO3 = 20 min 6.01 4.66 3.75 8.94 0.12 4.58

of the measured maximum intensities depicted in Figs. 2a–2c for various linear polar-
izer orientations and deposition times.
The phase retardations introduced by the inserted samples for deposition times of
10, 15, and 20 min were found to be 1.08, 0.79, and 1.33 rad, respectively. To determine
the appropriate m values, the maximal and minimal T values were tracked along with
the phase retardation. The thin-film thickness can be calculated using Eq. (19), e.g.,
d1(10 min) = (1.08 + 4π) λ /4 π n1 = 327.2 nm, d2(15 min) = (0.79 + 6π) λ /4 π n1 =
= 470.9 nm, and d3(20 min) = (1.33 + 8π) λ /4 π n1 = 634.5 nm. The measured thin-film
thicknesses using the CI system for deposition times of 10, 15, and 20 min were found
to be 327.2, 470.9, and 634.5 nm, respectively.

4.2. Spectroscopic ellipsometry (SE) and FE-SEM analysis

The values obtained from the VASE for the WO3 thin-film and the Tauc–Lorentz oscil-
lator [25] were used to plot the fitting curves of the ellipsometric parameters for different
deposition times. The VASE results showed that the WO3 thin-film-sample thicknesses
were 298.9, 440.3, and 595.5 nm for deposition times of 10, 15, and 20 min, respec-
tively. The samples were then placed under the field-emission scanning electron mi-
croscopy (FE-SEM) system to verify the results.
Figure 3 shows the cross-sectional FE-SEM micrographs of the WO3 thin-films on
the BK7 substrate for various deposition times. The films are compact, homogeneous,
and perfectly adherent to the BK7 substrate. The average thicknesses obtained from the

a b c

Fig. 3. Cross-sectional field-emission scanning electron microscopy (FE-SEM) images of WO3 thin-film
-sample thicknesses for deposition times of 10 min (a), 15 min (b), and 20 min (c).
Utilization of the cyclic interferometer in polarization phase-shifting technique... 79

image analysis of the cross-section sample for deposition times of 10, 15, and 20 min
were 313.6, 465.3, and 615.1 nm, respectively.

4.3. Result comparison and discussion

The film thickness measurements achieved with CI, SE, and FE-SEM were compared.
The thickness values obtained using the CI with SE and CI with FE-SEM exhibited
reading differences of 7.31%/3.15%, 6.95%/1.25%, and 6.54%/2.17% with respect to
the average readings from the standard tools for deposition times of 10, 15, and 20 min,
respectively. The longer the deposition time, the lower was the reading error. This could
be explained by the deposition rate being unsteady at the beginning and becoming more
linear with the passage of time. Nevertheless, the results showed no substantial devi-
ation between the SE, FE-SEM, and CI methods.

5. Conclusions
This study reports a new scheme to determine the thin-film thickness of a transparent
sample using a CI system. The dual RPL produced by our CI setup possessed an excellent
degree of polarization. Samples with smooth WO3 thin-films were prepared by using
magnetron sputtering. The films exhibited low absorption at high deposition rates over
deposition times of 10, 15, and 20 min. To verify the performance of the CI setup, the
experimental results obtained with our CI system were compared with the ones obtained
with standard VASE and FE-SEM techniques. There was no significant difference be-
tween the three sets of results. With our approach, it is possible to produce suitable
dual RPL, accurately investigate thin-film thickness, and characterize the phase retar-
dation of optical devices such as polarizers, wave retarders, and liquid-crystal variable
wave retarders.
Acknowledgements – The authors would like to acknowledge the contribution of the Thai Government
Science and Technology Scholarship for funding the project. The authors would also like to thank Dr. Mati
Horprathum for his time and advice, providing film samples, and performing film measurements.

References
[1] CHAKRABORTY S., BHATTACHARYA K., Low-level birefringence measurement by cyclic-path polariza-
tion interferometer, Applied Optics 55(21), 2016, pp. 5634–5639, DOI: 10.1364/AO.55.005634.
[2] FLORES MUÑOZ V.H., TOTO-ARELLANO N.I., LÓPEZ-ORTIZ B., MARTÍNEZ GARCÍA A., RODRÍGUEZ-ZURITA G.,
Measurement of red blood cell characteristic using parallel phase shifting interferometry, Optik
126(24), 2015, pp. 5307–5309, DOI: 10.1016/j.ijleo.2015.09.019.
[3] KUMAR Y.P., CHATTERJEE S., Thickness measurement of transparent glass plates using a lateral shear-
ing cyclic path optical configuration setup and polarization phase shifting interferometry, Applied
Optics 49(33), 2010, pp. 6552–6557, DOI: 10.1364/AO.49.006552.
[4] ROTHAU S., KELLERMANN C., MAYER S., MANTEL K., LINDLEIN N., Polarization and phase-shifting
interferometry for arbitrary, locally varying polarization states, Applied Optics 56(5), 2017, pp. 1422
–1430, DOI: 10.1364/AO.56.001422.
80 R. KAEWON et al.

[5] MALACARA D., Phase shifting interferometry, Revista Mexicana de Física 36(1), 1990, pp. 6–22.
[6] MENESES-FABIAN C, RIVERA-ORTEGA U., Phase-shifting interferometry by wave amplitude modula-
tion, Optics Letters 36(13), 2011, pp. 2417–2419, DOI: 10.1364/OL.36.002417.
[7] INDEBETOUW G., KLYSUBUN P., Measurement of induced birefringence in film samples using a bal-
anced polarization ring interferometer, Optics Communications 151(4–6), 1998, pp. 203–206, DOI:
10.1016/S0030-4018(98)00087-X.
[8] BHADURI B., KRISHNA MOHAN N., KOTHIYAL M.P., Cyclic-path digital speckle shear pattern interfer-
ometer: use of polarization phase-shifting method, Optical Engineering 45(10), 2006, article 105604,
DOI: 10.1117/1.2361194.
[9] SHAGAM R.N., WYANT J.C., Optical frequency shifter for heterodyne interferometers using multiple
rotating polarization retarders, Applied Optics 17(19), 1978, pp. 3034–3035, DOI: 10.1364/AO.17.
003034.
[10] KOTHIYAL M.P., DELISLE C., Shearing interferometer for phase shifting interferometry with polari-
zation phase shifter, Applied Optics 24(24), 1985, pp. 4439–4442, DOI: 10.1364/AO.24.004439.
[11] TOTO-ARELLANO N.-I., SERRANO-GARCÍA D.-I., MARTÍNEZ-GARCÍA A., Parallel two-step phase shifting
interferometry using a double cyclic shear interferometer, Optics Express 21(26), 2013, pp. 31983
–31989, DOI: 10.1364/OE.21.031983.
[12] ZHANG Z., ZHANG Q., CHENG T., GAO J., WU X., Wave-plate phase shifting method, Optical Engi-
neering 52(10), 2013, article 103109, DOI: 10.1117/1.OE.52.10.103109.
[13] SUJA HELEN S., KOTHIYAL M.P., SIROHI R.S., Achromatic phase shifting by a rotating polarizer,
Optics Communications 154(5–6), 1998, pp. 249–254, DOI: 10.1016/S0030-4018(98)00292-2.
[14] LECHUGA L.G., TOTO-ARELLANO N.I., FLORES MUÑOZ V.H., MARTÍNEZ-GARCÍA A., ZURITA G.R.,
Phase shifting interferometry using a coupled cyclic path interferometers, [In] Emerging Challenges
for Experimental Mechanics in Energy and Environmental Applications, Proceedings of the 5th In-
ternational Symposium on Experimental Mechanics and 9th Symposium on Optics in Industry
(ISEM-SOI), 2015, Martínez-García A., Furlong C., Barrientos B., Pryputniewicz R. [Eds.],Confer-
ence Proceedings of the Society for Experimental Mechanics Series, Springer, Cham, 2017, pp. 65–69,
DOI: 10.1007/978-3-319-28513-9_9.
[15] WANG L., LIU L., LUAN Z., SUN J., ZHOU Y., Polarization phase-shifting Jamin shearing interferom-
eter, Optik 121(4), 2010, pp. 358–361, DOI: 10.1016/j.ijleo.2008.07.023.
[16] CHAKRABORTY S., BHATTACHARYA K., Real-time edge detection by cyclic-path polarization interfer-
ometer, Applied Optics 53(4), 2014, pp. 727–730, DOI: 10.1364/AO.53.000727.
[17] LIU X., GAO Y., CHANG M., A new lateral shearing interferometer for precision surface measurement,
Optics and Lasers in Engineering 47(9), 2009, pp. 926–934, DOI: 10.1016/j.optlaseng.2009.03.019.
[18] PAWONG C., CHITAREE R., SOANKWAN C., The rotating linearly polarized light from a polarizing
Mach–Zehnder interferometer: production and applications, Optics and Laser Technology 43(3),
2011, pp. 461–468, DOI: 10.1016/j.optlastec.2010.06.020.
[19] KAEWON R., PAWONG C., CHITAREE R., BHATRANAND A., Polarization phase-shifting technique for
the determination of a transparent thin film’s thickness using a modified Sagnac interferometer, Cur-
rent Optics and Photonics 2(5), 2018, pp. 474–481.
[20] TATAM R.P., JONES J.D.C., JACKSON D.A., Optical polarization state control schemes using fibre op-
tics or Bragg cells, Journal of Physics E: Scientific Instruments 19(9), 1986, pp. 711–717, DOI:
10.1088/0022-3735/19/9/013.
[21] SARKAR S., BHATTACHARYA K., Polarization phase shifting cyclic interferometer for surface pro-
filometry of non-birefringent phase samples, Journal of Modern Optics 60(3), 2013, pp. 185–189,
DOI: 10.1080/09500340.2013.765051.
[22] MANIFACIER J.C., GASIOT J., FILLARD J.P., A simple method for the determination of the optical con-
stants n, k and the thickness of a weakly absorbing thin film, Journal of Physics E: Scientific Instru-
ments 9(11), 1976, pp. 1002–1004, DOI: 10.1088/0022-3735/9/11/032.
Utilization of the cyclic interferometer in polarization phase-shifting technique... 81

[23] SUN X., LIU Z., CAO H., Effects of film density on electrochromic tungsten oxide thin films deposited
by reactive dc-pulsed magnetron sputtering, Journal of Alloys and Compounds 504, 2010, pp. s418
–s421, DOI: 10.1016/j.jallcom.2010.03.155.
[24] ZHAO C., TAN J., TANG J., LIU T., LIU J., Confocal simultaneous phase-shifting interferometry, Ap-
plied Optics 50(5), 2011, pp. 655–661, DOI: 10.1364/AO.50.000655.
[25] FRANKE E., TRIMBLE C.L., DEVRIES M.J., WOOLLAM J.A., SCHUBERT M., FROST F., Dielectric function
of amorphous tantalum oxide from the far infrared to the deep ultraviolet spectral region measured
by spectroscopic ellipsometry, Journal of Applied Physics 88(9), 2000, pp. 5166–5174, DOI:
10.1063/1.1313784.

Received December 17, 2018