Miniature surface-mountable Fabry–Perot pressure
sensor constructed with a 45° angled fiber
H. Bae,1X. M. Zhang,2H. Liu,1and M. Yu1,*
1Department of Mechanical Engineering, University of Maryland, College Park, Maryland 20742, USA
2Department of Applied Physics, Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China
*Corresponding author: email@example.com
Received February 17, 2010; revised April 17, 2010; accepted April 20, 2010;
posted April 23, 2010 (Doc. ID 124310); published May 14, 2010
We present a surface-mountable miniature Fabry–Perot (FP) pressure sensor that utilizes the total internal reflection
at a 45° angled fiber end face to steer the optical axis by 90°. By using the fiber as a waveguide, as well as a natural
mask in photolithography, an FP cavity is constructed on the sidewall of the fiber. A polymer–metal composite
diaphragm is employed as the pressure transducer. The sensor exhibits a good linearity over the pressure range
of 1:9–14:2 psi, with a sensitivity of 0:009 μm=psi and a hysteresis of 2.7%. This sensor is expected to impact many
fronts that require reliable static pressure measurements of fluids.
060.2370, 120.2230, 230.3990.
© 2010 Optical Society of America
Miniature fiber-optic pressure sensors have become at-
tractive choices for pressure monitoring in a limited
space owing to their advantages of small size, high sen-
sitivity, immunity to electromagnetic interference, and
convenience of light guiding/detection through optical fi-
bers . Such a sensor typically exploits an extrinsic
Fabry–Perot (FP) interferometer formed directly on a
fiber end face, as illustrated in Fig. 1(a), which consists
of a cleaved optical fiber, a pressure sensing diaphragm,
and a housing structure for holding the diaphragm. The
fiber end facet and the reflective diaphragm effectively
form an FP cavity for sensing the external pressure.
As the FP cavity shares the optical axis with the optical
fiber, this configuration is named as coaxial configura-
tion. A variety of miniature coaxial FP sensors have been
reported in literature [2–9], and they have been applied to
many applications, including biomedical [6,8,9], aerody-
namic , and other industrial applications  where a
minimal intrusiveness is required.
However, for pressure measurements in fluids, a coax-
ial sensor that is positioned toward the flow can pick up
both dynamic and static pressures, which is known as the
total pressure . In this case, the flow is stopped at the
pressure-sensing diaphragm, resulting in an increase in
the pressure reading from the sensor. To eliminate the
dynamic pressure effect and to obtain only the static
pressure, a coaxial sensor should be carefully positioned
so that the axis of the FP cavity is perpendicular to the
direction of the flow. To facilitate such positioning,
mounting the coaxial sensor on a structure surface usual-
ly requires bending of the fiber and/or drilling a through-
hole on the surface, which is inconvenient and some-
times even impractical. Therefore, when it is important
to distinguish the static pressure from the surface flow,
an alternative sensor design that renders easy sensor
surface mounting is needed. Such surface-mountable
sensors are especially desirable for pressure measure-
ments in blood vessels [6,11] and on-blade pressure mon-
itoring of turbomachineries [3,7] and rotorcrafts .
In this Letter, we report a surface-mountable miniature
FP pressure sensor. The schematic of the sensor is
shown in Fig. 1(b), and it consists of a fiber with a 45°
angled end face, a sensing diaphragm, and a housing
structure formed on the sidewall of the fiber to hold
the diaphragm. The key concept here is to steer the op-
tical axis by 90° based on total internal reflection at the
45° angled fiber end face. In this configuration, the fiber
sidewall serves as a partial mirror of the FP interfero-
meter, and the reflective diaphragm that is parallel to
the fiber-optical axis serves as the other mirror. Because
the fiber axis is perpendicular to that of the FP cavity,
this configuration is called a cross-axial configuration.
One unique feature of the cross-axial sensor is that it can
be directly mounted on a structure surface or embedded
in a shallow channel to accurately measure the static
pressures even in the presence of surface flows.
The detailed fabrication process of the cross-axial FP
pressure sensor is shown in Fig. 2. First, preparation of
the angled fiber samples is performed [Fig. 2(a)]. A batch
(typically 20–50) of single-mode fibers (SMF28, Corning)
are spliced and glued on a glass slide by using a wax
ishing purposes. The fibers are polished to form the 45°
angled end face with a fiber polishing machine. Calibra-
tion of the polished angles is carried out by measuring
the reflection fromthepolishedfiberendface. Thereflec-
A thin layer of silver is then evaporated onto the angled
fiber end face to further enhance the reflection.
Second, a photoresist layer is deposited and UV ex-
posed; this serves as the housing structure. The polished
fiber is dipped into a positive photoresist (AZ 4620, Ship-
ley) to cover the fiber sidewall with a thin layer of photo-
resist. Soft baking is then carried out at 95 °C for 5 min.
pressure sensors: (a) coaxial configuration and (b) cross-axial
(Color online) Configurations of miniature fiber-optic
May 15, 2010 / Vol. 35, No. 10 / OPTICS LETTERS 1701
0146-9592/10/101701-03$15.00/0© 2010 Optical Society of America
After each cycle of deposition and soft baking, a photo-
resist layer of 3–5 μm will be formed on the fiber side-
wall. As the cavity length of the FP interferometer is
determined by the photoresist thickness, which is mon-
itored by using an optical microscope (Olympus) and a
digital camera (CFW-1308C, Scion Corporation), various
cavity lengths can be obtained by performing different
numbers of deposition/soft-baking cycles. After a desired
thickness is obtained, light from a spot UV source (Blue-
Wave 50S, DYMAX) is coupled into the fiber and re-
flected by the 45° angled end face to expose the
photoresist along the light path; see Fig. 2(b).
Third, an air cavity serving as the FP cavity is formed in
the photoresist. The exposed photoresist is developed by
using a developer (AZ 440 K diluted 1∶3) for 5 min to
form the air cavity, as shown in Fig. 2(c). The sample is
then hard baked at 100 °C for 5 min to enhance the sta-
bility of the remaining photoresist housing structure. The
depth of the cavity is measured by using an optical profil-
ometer (TMS 1200, Polytec) with a high resolution
(0:195 nm). As the fiber waveguide is used as a photo
accurately without the need for any alignment systems or
extra masks .
Finally, a polymer–metal composite diaphragm is pre-
pared and covered on the housing. A drop of acrylic ur-
ethane UV polymer (OP-4-20641, DYMAX) is dispensed
into deionized water in a petri dish (diameter of
100 mm). After spreading to a thin layer with a desired
thickness, the polymer is half cured, and then lifted up
and covered onto the housing structure. After full curing,
a silver layer with a chromium adhesion layer is sput-
tered on the polymer film to enhance the reflectivity. An-
other optional protective polymer layer can be added
later to isolate the metal layer from the external environ-
ment to avoid oxidization and corrosion. A detailed dia-
phragm preparation process can be found in . The
completed surface-mountable FP sensor is illustrated in
Fig. 2(d). Both the thicknesses of the metal layer and the
polymer layer can be tuned to meet the specific needs of
various pressure-sensing applications.
In this work, a 1:3 μm acrylic urethane film coated with
150 nm silver and 3 nm chromium was used as the dia-
phragm and a 13:8 μm depth optical cavity was obtained
on the fiber sidewall. The overall sensor diameter
(∼150 μm) is slightly larger than the fiber diameter
(125 μm). The diaphragm diameter was measured to
be around 18 μm. The pressure sensitivity in terms of
the diaphragm center deflection is predicted to be
0:0131 μm=psi by using a finite element model. A scan-
ning electron micrograph of the cavity before applying
the diaphragm and a close-up of the composite dia-
phragm are shown in Fig. 3.
In the experiment, the fabricated FP sensor was con-
nected to a white light interferometer system for interro-
gation. The system consists of a broadband light source
(HL-2000 Tungsten Halogen Light Source, Ocean Optics),
a 2 × 2 coupler (50=50), a spectrometer (USB4000, Ocean
Optics), and a computer for data collection and sig-
nal processing . The absolute cavity length L [see
Fig. 1(b)] can be retrieved from the reflection spectrum
of the spectrometer as L ¼
two adjacent peaks of the reflection spectrum and FSR ¼
jλ2− λ1j is the free spectral range . The spectrometer
has a spectral resolution of ∼0:1 nm, rendering a cavity
length resolutionof 3:5 nm. Toreduce the error in finding
the FSR, the curve fitting method was used . The cav-
ity length L was eventually calculated by averaging sev-
eral FSRs obtained from the peaks with good visibility to
further reduce the random errors.
Sensor calibration was conducted in a pressure cham-
ber with a reference pressure sensor (LL-080-25A, Kulite
Semiconductor Products). The pressure in the chamber
was controlled by using a pressure regulator (R-68825-08,
Marsh Bellofram) with a pressure regulation range of 1.9
to 60 psi. In the experiment, the pressure was first in-
creased from 1.9 to 14:2 psi at room temperature, and
then it was gradually decreased to atmospheric pressure.
Figure 4 shows the calibration results, which exhibit a
good linearity (R2¼ 0:99). Based on the linear fitting
of the experimental data, the pressure sensitivity of
the sensor can be obtained to be around 0:009 μm=
psi in the pressure range of 1:9–14:2 psi, which com-
pareswell with themodel
(0:0131 μm=psi). Since the calibration curves obtained
from increasing and decreasing pressures are almost
overlapped, the hysteresis error of the sensor is not sig-
nificant (∼2:7%). It is noted that the results exhibit a drift
of the initial cavity length by 0:005 μm, which is about
0.04% of the initial cavity length (13:8 μm). This small
drift is found to be due to the ambient temperature var-
iation (around a couple of degrees) observed during the
2FSR, where λ1and λ2are any
mountable sensor: (a) polishing of the 45° angled fiber end face
and deposition of the metal layer, (b) exposure of the photore-
sist housing with a fiber-coupled UV light, (c) development of
the photoresist, and (d) covering of the polymer diaphragm and
deposition of the reflective layer.
(Color online) Fabrication process flow of the surface-
fore covering the diaphragm and (b) the close-up of the com-
Scanning electron micrographs of (a) optical cavity be-
1702OPTICS LETTERS / Vol. 35, No. 10 / May 15, 2010
experiment. Over a temperature range of T ¼ 28 °C to Download full-text
T ¼ 40 °C, the thermal effect on the sensor has been
carefully studied. As shown in the inset of Fig. 4, both
the initial cavity length and the pressure sensitivity show
slight temperature-associated variations. In applications
where temperature variations are of concern, an external
temperature sensor or an additional FP cavity with a dif-
ferent initial cavity length can be included to provide
To the best of the authors’ knowledge, this is the first
time that a miniature cross-axial FP pressure sensor has
been developed. The cross-axial FP sensor possesses
several unique features in terms of sensor design and fab-
rication as compared to the previously reported coaxial
FP pressure sensors. First, the FP cavity is achieved on
the fiber sidewall, rendering a surface-mountable sensor
that can be simply attached to the structure surface for
reliable static pressure measurements. Second, the fabri-
cation follows simple, repeatable processes and safe pro-
cedures, and uses less expensive materials, without the
need for a clean-room environment. Although photolitho-
graphy is utilized for patterning the photoresist housing
structure, the fabrication does not need any extra photo
masks, as the FP optical cavity is self-aligned to the
steered light path. This eliminates tedious optical align-
ments, which are usually quite challenging in the fabrica-
tion of a miniature fiber-optical sensor. Third, the
multilayer design of the sensor diaphragm not only en-
hances the durability of the sensor but also allows for
easy tuning of the sensor performance parameters
(i.e., sensitivity and dynamic range).
In conclusion, a surface-mountable miniature FP pres-
sure sensor is developed, which takes the advantage of
the total internal reflection occurred at a 45° angled fiber
end face to facilitate both cross-axial pressure sensing
and photolithography. The experiment shows that the
sensor has a linear response over a pressure range of
1:9–14:2 psi, with a sensitivity of 0:009 μm=psi. It is ex-
pected that this sensor will benefit many areas that re-
quire miniature sensors for reliable static pressure
measurements of fluids and gases.
We gratefully acknowledge support from the National
Science Foundation (NSF) (CMMI0644914), the Center
for Rotorcraft Innovations (W911W6-06-2-0002), the Uni-
ted States Army Research Office (USARO) Defense Uni-
versity Research Instrumentation Program (DURIP)
(W911NF0710215), and the Maryland NanoCenter and
1. Y. J. Rao, Opt. Fiber Technol. 12, 227 (2006).
2. S. Watson, M. J. Gander, W. N. MacPherson, J. S. Barton,
J. D. C. Jones, T. Klotzbuecher, T. Braune, J. Ott, and F.
Schmitz, Appl. Opt. 45, 5590 (2006).
3. W. N. MacPherson, J. M. Kilpatrick, J. S. Barton, and
J. D. C. Jones, Rev. Sci. Instrum. 70, 1868 (1999).
4. X. W. Wang, J. C. Xu, Y. Z. Zhu, K. L. Cooper, and A. Wang,
Opt. Lett. 31, 885 (2006).
5. D. C. Abeysinghe, S. Dasgupta, J. T. Boyd, and H. E.
Jackson, IEEE Photonics Technol. Lett. 13, 993 (2001).
6. K. Totsu, Y. Haga, and M. Esashi, J. Micromech. Microeng.
15, 71 (2005).
7. M. J. Gander, W. N. MacPherson, J. S. Barton, R. L. Reuben,
J. D. C. Jones, R. Stevens, K. S. Chana, S. J. Anderson, and
T. V. Jones, IEEE Sens. J. 3, 102 (2003).
8. E. S. Olson, J. Acoust. Soc. Am. 103, 3445 (1998).
9. S. Nesson, M. Yu, X. M. Zhang, and A. H. Hsieh, J Biomed.
Opt. 13, 044040 (2008).
10. V. L. Streeter and E. B. Wylie, Fluid Mechanics, 8th ed.
11. R. Melamud, A. A. Davenport, G. C. Hill, I. H. Chan, F.
Declerq, P. G. Hartwell, and B. L. Pruitt, in Proceedings
of 18th International Conference on Microelectromechani-
cal Systems (2005), pp. 810–813.
12. Y. Liu, L. Alexander, G. Wang, P. Ashish, and M. Yu, Proc.
SPIE 6770, 67700Y (2007).
13. M.-S. Kim, K.-W. Jo, and J.-H. Lee, Appl. Opt. 44, 3985
14. K. T. V. Grattan and B. T. Meggitt, Optical Fiber Sensor
Technology (Chapman & Hall, 1995).
15. B. Qi, G. R. Pickrell, J. Xu, P. Zhang, Y. Duan, W. Peng,
Z. Huang, W. Huo, H. Xiao, R. G. May, and A. Wang, Opt.
Eng. 42, 3165 (2003).
(Color online) Sensor cavity length as a function of the
May 15, 2010 / Vol. 35, No. 10 / OPTICS LETTERS1703