Grating surface plasmon resonance sensor: angular

sensitivity, metal oxidization effect of Al-based

device in optimal structure

Xiaoliang Sun, Xuewen Shu, and Changhong Chen*

Wuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology,
Wuhan 430074, China
*Corresponding author: ch_chen@hust.edu.cn

Received 6 October 2014; revised 13 January 2015; accepted 21 January 2015;

posted 22 January 2015 (Doc. ID 224396); published 20 February 2015

We present the wavelength (λ)-independent ultimate and maximum angular sensitivity of a grating sur-
face plasmon resonance sensor, as long as the grating of metal has much higher permittivity than analyte
(n2a ). For 1.32 ≤ na ≤ 1.36, the maximum angular sensitivity reaches 493.7 (535.9)°/RIU by the single
(double)-dip method. The real Al-based sensor in an optimal structure for a high figure of merit exhibits
higher sensitivity but narrower reflection dip when working at a longer wavelength, and the real sensi-
tivity increases from 292.5 (344.5) to 338.0 (396.3)°/RIU at respective λ
0.85 and 1.55 μm before the
metal oxidation. After the oxidization, the sensitivity is degraded by ≤3.2% (2.8%), and the detection
error of δna ≤ 1.3e − 3 (1.1e − 3) is introduced to the sensor working at near-infrared wavelengths of
interest. © 2015 Optical Society of America
OCIS codes: (050.6624) Subwavelength structures; (240.6680) Surface plasmons; (280.4788) Optical
sensing and sensors.
http://dx.doi.org/10.1364/AO.54.001548

1. Introduction Recently, many researchers have aimed at improv-

Since 1982, when surface plasmon resonance (SPR) ing the angular sensitivity of a metal grating SPR
was first introduced to detect small changes in re- sensor to detect smaller changes in the refractive in-
fractive index at a metal-analyte interface [1], SPR dex of analyte (na ) [11,12]. For example, in contrast
sensing technology has received continuous atten- with the single-dip method (SDM), the sensitivity
tion because it takes advantage of a high-sensitivity, can be improved if the double-dip method (DDM)
real-time, label-free, and fast-speed response [2–4]. is applied, and, in a range of na
1.32–1.37, an aver-
Conventional SPR sensors are based on the attenu- aged real value of Sreal
237.0°∕RIU (per refractive
ated total reflection method in a prism-based struc- index of unit) was achieved when the sensor works
ture to excite the surface plasmon wave (SPW); at the wavelength of λ
0.85 μm. We, therefore, ask
however, the prism-based SPR sensors need precise what ultimate sensitivities (or maximum values when
angle control and are too bulky for integration [5]. As na changes in a measuring range) can be achieved?
an alternative, an SPR sensor based on a binary Are they dependent on the wavelength and the gra-
metal grating has been proposed by Cullen et al. [6]; ting geometrical structure and material? So far, we
since then, grating-based SPR sensors have been still lack the detailed analysis for the ultimate or
widely studied [7–10]. maximum values.
On the other hand, as one of the SPR-active met-
als, aluminum (Al) has been proven to provide
sharper SPR peaks than Ag, Au, or Cu; [13] however,
1559-128X/15/061548-07$15.00/0 if Al is selected, the metal grating will oxidize to
© 2015 Optical Society of America change its structure no matter where it is in air or

1548 APPLIED OPTICS / Vol. 54, No. 6 / 20 February 2015

in most of the analytes. Generally, an approach to Al [14]. Owing to poor chemical stability, the metal in
protecting the Al layer from oxidizing is to design the air or aqueous solution will oxidize to Al2 O3 coat-
the grating in a bimetallic-layer structure of Au/Al, ing on the grating surface, as shown in Fig. 1(a). The
[14–17]; however, it causes disadvantages of not only oxide is compact enough to prevent oxygen atoms
degrading the sensitivity and accuracy but raising from further diffusing, so the oxidation is finally
the complexity of the production process. Thus, we suppressed to saturate its depth at Δh ∼ 4.5 nm [18].
want to know, if the Al grating in a single-layer struc- The metal layer of the grating is designed thick
ture is retained, what is the oxidization effect of the enough to neglect impact from the substrate.
metal on the sensor performance? Will the sensitiv- Figure 1(b) shows dispersion relations of the SPW
ities degrade to introduce significant error for detec- and the incident light wave. When TM-polarized
tion of the analyte index? Unfortunately, the answers light of wavelength (λ) is incident on the grating sur-
are not yet quite clear. face, the grating diffraction provides excess momen-
Therefore, it is meaningful to determine the ulti- tum, Δκ
m · 2π∕p, to compensate the wave vector
mate or maximum sensitivities and to clarify the mismatch between κlight and κSPW , and the SPW is,
oxidization effect on the performance of an Al-grating thus, excited. To probe the reflecting light in the
SPR sensor by the methods when it works at differ- incident angle (θ) scan, we can observe reflection
ent near-infrared wavelengths of interest, e.g., dip, where the amplitude is sharply attenuated at
λ
0.85, 0.98, 1.31, and 1.55 μm. Here, we present θ
θr;m , as the SPW is highly concentrated on the
expressions of the ultimate sensitivity and answer grating surface. If θ > θr;m, the concentration be-
all the questions one by one. comes weaker at a larger angle of incidence; while
θ < θr;m , no SPW is excited for greater wave vector
2. Theory of Grating-based SPR Sensor departure from the light line, and the incident light
SPW is a special form of an electromagnetic wave is scattered.
launched from the collective oscillation of the surface
charge density at a metal-dielectric interface. In the 3. Results and Discussion
grating-based system, the match condition of the In the infrared spectral range, SPR offers high probe
wave vector is given by the expression: depth [19] and precise determination of reflection

 dips [20]. Moreover, jεM j ≫ n2a is true for Al (also for
2π εM n2a 1∕2 Ag, Au, Cu, etc.) if the refractive index of analyte
κ0 na sin θr;m  m
κ 0 ; (1)
p εM  n2a changes around na ∼ 1.32, and the resonant angle
of θr;m
sin−1 1 − mλ∕pna  is obtained. Here, the
where κ o is the free space wave vector of the optical grating period must be satisfied as
wave, θr;m is the incident angle in resonance with
minimal reflection of light, m is a nonzero integer 
p > m nλa ; if m > 0;
to represent the diffraction order, p is the grating jmj λ 4
period, and εM is the grating permittivity of metal. 2 na <p< jmj nλa ; if m < 0:
Signs “+” and “−” correspond to m > 0 and m < 0,
respectively. Generally, the effective excitation of the Figure 2 plots the resonant angles in different
SPW requires jεM j ≫ n2a . diffraction order. In general, negative orders can lead
An important parameter to evaluate the sensor to larger θr;m, which is a monotonic function of p. If
performance, angular sensitivity is defined as the λ∕2na  < p < λ∕na , only one mode exists in order
deviation of the resonant angle to the analyte index m
−1; while λ∕na < p < 3λ∕2na , two modes in
of refraction. The ultimate sensitivity for a sensor
working in multiple diffraction orders is expressed as
   
X dθr;m  X  Δθr;m 
Sult
   :
 dn 
lim
Δna →0 Δna 
(2)
m a m

Here, S̄m
jΔθr;m · Δn−1

a j is the average sensitivity,
dependence on the measuring range, for a single
mode.
Al is used as the active metal in this work, and the
complex permittivity is described by the following
lossy Drude model:

λ2 λc
εM λ
1 − ; (3) Fig. 1. (a) Schematic of Al-grating SPR sensor based on dielectric-
λ2p λc  iλ metal diffraction. (b) Dispersion relations of SPW and incident
wave. Grating parameters: p is the period, w0 w is the width of
where λ is the wavelength, λp is the plasma wave- metal, and h is the depth. Besides, θ is the incident angle of light
length, and λc is the collision wavelength. We have wave, Δh is the depth of oxide, and d is the metal thickness. κ light ,
λp
1.0657e − 1 μm and λc
2.4511e  1 μm for incident wave vector; κ SPW , SPW vector.

20 February 2015 / Vol. 54, No. 6 / APPLIED OPTICS 1549

 Fig. 3. For 1.32 ≤ na ≤ 1.36, grating period (p) dependence of res-
Fig. 2. Grating period (p) dependence of resonant angle θr;m  in onant angle (θr;−1 ) and ideal angular sensitivity (Sideal ).
diffraction orders m
1, 2, −3, and −4.

smaller, the overall surface area tends to be larger

orders m
−2 and 1 will be present. As p further to induce a notable amplification of the SPR signal
increases, more modes coexisting as a higher order of [21]. Considering the tolerance in fabrication, we
diffraction is possible. We deduce the sensitivity to select p
588 nm as the optimized period and ob-
each mode as Sm
j1 − sin θr;m na · cos θr;m −1 j. tain θr;−1
85.6° and 69.8° at na;0 and na;1 ; therefore,
If θr;m are the same angles for different modes, the
the ideal sensitivity up to Sideal
395.0°∕RIU is
diffraction in order m < 0 always creates a larger
feasible to the SDM. If the sensor works at other
sensitivity than m > 0; while, within an identical
near-infrared wavelengths of interest, λ
1.31,
mode, the smaller p is, the higher θr;m and, thus,
0.98, and 0.85 μm, the optimal grating period is
Sm are.
p
497, 372, and 322 nm, and the corresponding
We deal with an aqueous solution as the analyte, of
ideal sensitivity is determined as Sideal
392.5,
which the refractive index changes from na;0
1.32
377.5, and 467.5°/RIU.
to na;1
1.36. In order to keep the diffraction modes
We investigate the impact from the grating width
locked when na changes, the metal-grating period is
and depth on the resonance reflectance curves, based
satisfied in narrowed ranges of λ∕2na;0  < p < λ∕na;1
on rigorous coupled wave analysis [22], and use the
to the SDM, while λ∕na;0 < p < 3λ∕2na;1  to the DDM.
overall figure of merit (FOM) to evaluate the SPR
A. Single-Dip Method sensor performance. To the reflection dip of m-mode,
As the diffraction order m
−1 is applied only to the corresponding mode FOM is defined as [21]
SDM, we have the angular sensitivity, S
S̄m
−1 .
The ultimate sensitivity achieved at p
λ∕2na  is Sideal 1 − MRR
FOMm
; (6)
described as FWHM

 
na 1∕2 −1 where MRR is the reflectance at θr;m and FWHM is
Sult
lim S
lim 2 na  Δna  ; (5)
Δna →0 Δna →0 Δna the full width of the dip at half-maximum. The
deeper and narrower the dip is, the more accurate
which exhibits independence of the working wave- it is to detect the refractive index of analyte by the
length and the grating structure except for the shift of θr;m. Thus, FOM
FOMm
−1 for the SDM.
period, as long as jεM j ≫ n2a. When the refractive Before the oxidation, sets of the FOM value are
index changes in the measuring range from na;0 to obtained for the grating with various widths but con-
na;1 , a maximum value of Smax
493.7°∕RIU is ex- stant depth, which remains the same in oxidation,
pected; however, it decreases as na;1 increases. Smax when the refractive index changes from na;0 to na;1 .
is λ-independent, too. Among all the FOM values, we choose and plot the
Contrary to the real sensitivity, decided by Δna and maximum (FOMmax ) achieved at w
w0 in Fig. 4.
structural parameters and material of the grating, To clarify the oxidization effect, the grating in
we define the ideal sensitivity as a function of Al2 O3 ∕Al double layers is considered [see Fig. 1(a)].
p; Sideal
S̄m
−1 p, the average sensitivity not to The oxide has a refractive index of n
1.6, while its
consider has an impact from the grating width and dispersion effect can be ignored in the near-infrared
depth. If λ
1.55 μm, we obtain the grating period in spectral range [23]. The maximum depth of oxide is
a narrowed range of 587.1 < p ≤ 1139.7 nm. Figure 3 limited at Δh
4.5 nm. During the oxidation, the
shows the grating period dependence of the resonant grating width in part of the metal narrows to w0

angle and the ideal sensitivity. It indicates that θr;−1 w0 − 2Δh∕f V , while its period and depth will
and Sideal increase as p deceases; especially, the remain unchanged. Here, the volume factor of the
change is sharp around p
600 nm. When p is oxide to the metal is f V ∼ 1.3. [24] In contrast, we also

1550 APPLIED OPTICS / Vol. 54, No. 6 / 20 February 2015

Fig. 4. To the SDM, grating depth (h) dependence of FOM at Fig. 5. To the SDM, real angular SPR curves in reflectance (RTM )
different working wavelength: (a) λ
1.55; (b) 1.31; (c) 0.98; and for the optimized Al-grating SPR sensor working at λ
1.55 μm.
(d) 0.85 μm. FOM is plotted for optimizing the grating width
and depth. In the case of constant depth, it reaches FOMmax at enough to lower readout errors of the resonant angle.
w
w0 before the oxidation, while it is degraded for the device On the other hand, as shown in Fig. 6(a), the angular
in a narrowed metal grating width (w0 ) after the oxidation. Here, SPR curve with different oxide depth is used to dem-
w0 is searched by a step of 5 nm, and FOM is calculated by select-
ing na;0 as reference.
onstrate the shift of resonant angle in the oxidation.
During the oxidation, MRR and FWHM increase
slightly, whereas θr;−1 decreases as the oxide depth
plot the degraded FOM, owing to an influence of the increases. It is attributable to an increasing effective
metal grating width narrowed from w0 to w0 after the refractive index of the grating with formation and
metal oxidation. growth of the oxide layer. Figure 6(b) shows drifts of
Figures 4(a)–4(d) shows FOM of the device work- FWHM and θr;−1 resulting from the oxidation. After
ing at different wavelengths. Before the oxidation, the oxidation, larger drifts of FWHM and θr occur at
FOMmax can be achieved at the labeled width. If a lower na ; however, ΔFM ≤ 0.04°, i.e., a slight degra-
λ
1.55 μm, e.g., we determine w0 ∼ 320, 445, 565, dation of precision, and δθr;−1 in the same trend
or 560 nm when h
10–50 nm at 5 nm intervals. results in Δθr;−1 reduced only by 0.18° between na;0
FOMmax -peaks are observed at h ∼ 35, 27, 20, and and na;1 . The angular sensitivity degrades to S0real

17 nm for λ
1.55, 1.31, 0.98, and 0.85 μm. However, 333.5°∕RIU by the change of ∼1.3%, defined as
it is noteworthy that FOMmax continues to decrease 1 − S0real ∕Sreal , so an error of ∼5.3e − 4 will be intro-
as h further increases; besides, as shown in Figs. 4(a) duced to the refractive index. The detection precision
and 4(b), for the longer wavelength, another and sensitivity for the SPR sensor at λ
1.55 μm are
FOMmax -peaks will appear at the lower grating only changed insignificantly after the oxidation.
depth, where the width (w0 ) is generally much nar- If the Al-grating SPR sensor works at the other
rower. Since a large duty cycle (w0 ∕p) produces the wavelengths, the grating structure parameters and
high coupling efficiency from the incident optical the oxidization effect are investigated by the same
wave to the SPW, only the peak from a wider grating method to the device at λ
1.55 μm. Besides the
is considered. After the oxidation, degradation of the optimal structure and performance, Table 1 lists
FOM is obvious, and it can be interpreted as the sig-
nificant increase in the dip’s FWHM and MRR
caused by the narrowed w0 . On the other hand, deter-
mining the grating width and depth, we also consider
that, even after the metal oxidation, the device still
retains much higher FOM. Therefore, p; w; h

588; 565; 35, (497, 475, 30), (372, 350, 25), and (322,
305, 20); the parameter in each nanometer is an
optimized grating structure for λ
1.55, 1.31, 0.98,
and 0.85 μm, respectively.
We select the device working at λ
1.55 μm to
illustrate the determination of real sensitivity.
Figure 5 shows angular SPR curves at the different
na for the optimal structure. It indicates that the
resonant angle decreases as na increases, but reflec-
tion amplitude (RTM ) increases gradually. We obtain Fig. 6. For the optimized sensor: (a) during the oxidation, real
θr;−1
82.16° and 68.64° at na;0 and na;1 , so a real angular SPR curves in reflectance (RTM ) at na;1 ; (b) after the
angular sensitivity of Sreal
338.0°∕RIU is achieved. oxidation, drifts of FWHM and resonant angle, ΔFW and −δθr;−1 ,
Averaged dip width of FWHM
0.32° is small at different na · Δh
4.5 nm is the saturated oxide depth.

20 February 2015 / Vol. 54, No. 6 / APPLIED OPTICS 1551

 Table 1. To the SDM, Optimal Grating Structure of Period (p), Width (w ), and Depth (h)a

Grating Structure Performance Error

λ (μm) p (nm) w (nm) h (nm) Sreal (°/RIU) S0real (°/RIU) FWHM (°) FWHM0 (°) δna
0.85 322 305 20 292.5 283.0 0.51 0.57 1.3e − 3
0.98 372 350 25 292.5 287.3 0.49 0.60 7.1e − 4
1.31 497 475 30 324.5 319.8 0.35 0.39 5.8e − 4
1.55 588 565 35 338.0 333.5 0.32 0.34 5.3e − 4
a
Real sensitivity of Sreal and S0real , averaged dip width of FWHM and FWHM0 , and introduced detection error of δna are listed when
the sensor works at the different wavelength.

the detecting error of na from the oxidization effect The grating width and depth are optimized to
for the other wavelengths of interest. It reveals that realize high FOM for the DDM, too. It is the sum
the device working at a longer wavelength generally of the two mode figures of merit, FOM
FOMm
1 
exhibits the higher angular sensitivity and narrower FOMm
−2 . Figure 8 illustrates the grating depth
reflection dip. From the metal oxidation, the de- dependence of FOM for the device working at the
graded sensitivity of 283.0°/RIU is the lowest at different wavelengths. Before the oxidation, FOM
λ
0.85 μm, while the highest sensitivity and the changes in a similar trend to the SDM when the
narrowest FWHM are reached at λ
1.55 μm. More- device works at the identical wavelength, and an
over, the sensitivities are degraded by ≤3.2%, and the
introduced detection errors of na are not higher than
1.3e − 3; so the oxidization effect is considered to be
negligible to the SDM.
B. Double-Dip Method
As two diffraction orders, m
1 and −2, are applied
to the DDM, the angular sensitivity, S
S̄m
1 
S̄m
−2 , is a sum of the sensitivity for the different
modes:
   
 Δθr;1   Δθr;−2 

S
    : (7)
Δna   Δna 

Here θr;1 and θr;−2 correspond to the resonant angles

in the simultaneously existing modes. The ultimate Fig. 7. For 1.32 ≤ na ≤ 1.36, grating period (p) dependence of
sensitivity will occur at p
λ∕na , which is described as resonant angle (θr;1 ; θr;−2 ) and ideal angular sensitivity (Sideal ).

  
na 1∕2 −1
Sult
lim S
lim 12 na  Δna  :
Δna →0 Δna →0 Δna
(8)

It is higher than that to the SDM and also exhibits

independence of the working wavelength and the
grating structure, except for the period, as long as
jεM j ≫ n2a . A λ-independent maximum sensitivity of
Smax
535.9°∕RIU can be expected in the range from
na;0 to na;1 , but it decreases with an increase of na;1.
To λ
1.55 μm, the grating period falls in the
narrowed range of 1174.2 < p ≤ 1709.5 nm. Figure 7
shows the resonant angles and the ideal sensitivity
dependence of the grating period. As p increases,
θr;1 increases, while θr;−2 and, thus, Sideal decrease.
We select p
1175 nm as the optimized period and Fig. 8. To the DDM, grating depth (h) dependence of FOM at
different working wavelength: (a) λ
1.55; (b) 1.31; (c) 0.98; and
obtain the ideal sensitivity of Sideal
468.3°∕RIU.
(d) 0.85 μm. FOM is plotted for optimizing the grating width
For the other wavelengths, the optimal grating and depth. In case of the constant depth, it reaches FOMmax at
periods of p
993, 743, and 644 nm are determined w
w0 before the oxidation, while is degraded for the device in
and correspond to the ideal sensitivities of 471.8, narrowed metal grating width (w0 ) after the oxidation. Also, w0
462.5, and 508.8°/RIU at λ
1.31, 0.98 and 0.85 μm, is searched by a step of 5 nm, and FOM is calculated by selecting
respectively. na;0 as reference.

1552 APPLIED OPTICS / Vol. 54, No. 6 / 20 February 2015

 −2 dips are calculated for the different na and drift
from 0.07° to 0.08° and from 0.18° to 0.19°, respec-
tively. Therefore, the degradations of sensitivity and
the precision from the metal oxidization are insignifi-
cant for the SPR sensor working at λ
1.55 μm.
A device that works at the different wavelength is
also investigated. Table 2 lists the respective optimal
structure, performance, and detecting error of na from
the oxidization effect. Working at the same wave-
length, the device to the DDM always has the higher
sensitivity than that to the SDM, while the difference
of detection error is slight between the methods.
Moreover, the device working at a longer wavelength
exhibits higher angular sensitivity and narrower
Fig. 9. (a) To the DDM, real angular SPR curves of reflectance
reflection dip on average. The degraded sensitivity
(RTM ) in diffraction orders m
1 and −2 for the optimized
Al-grating SPR sensor working at λ
1.55 μm. (b) Before and after
of 334.8°∕RIU is the lowest at λ
0.85 μm, while
the oxidation, the plots of MRR and resonant angle, θr;1 and θr;−2 , the highest sensitivity of 390.5°∕RIU is obtained at
at different na . λ
1.55 μm. As the degradation of sensitivity and
the introduced detection errors of na do not exceed
2.8% and 1.1e − 3, respectively, the oxidization effect
ultimate value of FOMmax can be obtained at a large is also negligible to the DDM.
grating width (w0 ). For the requirement of a larger
grating width and the high FOM to be retained after
the oxidation, we obtain p; w; h
1175; 1150; 30, 4. Conclusion
(993, 950, 30), (743, 720, 20), and (644, 620, 15), i.e., In summary, we have given the expressions of ulti-
the parameter in each nanometer is an optimized mate angular sensitivity for a grating-based SPR
grating structure for λ
1.55, 1.31, 0.98, and sensor, which exhibit independence of the working
0.85 μm, respectively. wavelength and grating structure parameters, ex-
We also select the device working at λ
1.55 μm cept for the period, as long as the metal grating has
to illustrate the determination of real sensitivity. much higher permittivity than the analyte. By the
Figure 9(a) shows real angular SPR curves at the single (double)-dip method, the maximum angular
different na for the optimal structure. In each curve, sensitivity also is λ-independent and reaches 493.7
reflection dip from the diffraction order m
1 is (535.9)°/RIU if na changes in the range from 1.32 to
always shallower than that from the diffraction 1.36. The binary Al-grating structure of the real
order m
−2. As na increases, the reflection ampli- sensor is optimized for the high FOM requirement,
tude (RTM ) decreases for m
1 but increases for and the real sensitivity increases from 292.5
m
−2 gradually. Changes of the resonant angle (344.5)°/RIU, for λ
0.85 μm to 338.0 (396.3)°/RIU
are Δθr;1
1.72° and Δθr;−2
−14.13° when na in- for λ
1.55 μm before the metal oxidation. However,
creases from na;0 to na;1 ; thus, the real sensitivity, after the oxidization, the sensitivity is degraded by
Sreal
396.3°∕RIU, is achieved to the DDM. ≤3.2% (2.8%), generally slighter at longer wave-
Figure 9(b) shows changes of the MRR and the res- length, and the error of δna ≤ 1.3e − 3 (1.1e − 3) is
onant angles resulting from the oxidation. After the introduced for na-detection. The sensor needs recali-
oxidation, the reflection dips for m
1 and m
−2 bration to eliminate the error only if detecting a
become shallow, and increases of θr;1 but decreases much smaller change of na, so the oxidization effect
of θr;−2 are also observable. However, the real sensi- is negligible for the sensor when working at the wave-
tivity of Sreal 0
390.5°∕RIU is degraded by a change lengths of interest in the near-infrared spectral range.
of <1.5%. It indicates that a na -error of δna ∼ 5.9e − 4 This work was supported by the National Natural
will be introduced. Besides, FWHM of m
1 and Science Foundation of China, grant no. 61176064.

Table 2. To the DDM, Optimal Grating Structure of Period (p), Width (w ), and Depth (h)a.

Performance
Grating Structure FWHM (°) FWHM0 (°)
Sreal S0real
λ (μm) p (nm) w (nm) h (nm) (°/RIU) (°/RIU) m
1 m
−2 m
1 m
−2 Error δna
0.85 644 620 15 344.5 334.8 0.11 0.38 0.12 0.43 1.1e − 3
0.98 743 720 20 352.0 343.5 0.10 0.38 0.12 0.40 9.6e − 4
1.31 993 950 30 386.8 384.5 0.09 0.37 0.10 0.42 2.4e − 4
1.55 1175 1150 30 396.3 390.5 0.07 0.18 0.08 0.19 5.9e − 4
a
Real sensitivity of Sreal and S0real , averaged dip width of FWHM and FWHM0 , and introduced detection error of δna are listed when
the sensor works at the different wavelength.

20 February 2015 / Vol. 54, No. 6 / APPLIED OPTICS 1553

References 13. A. K. Sharma and B. D. Gupta, “On the performance of differ-
ent bimetal combinations in surface plasmon resonance based
1. C. Nylander, B. Liedberg, and T. Lind, “Gas detection by fiber optic sensor,” J. Appl. Phys. 101, 093111 (2007).
means of surface plasmon resonance,” Sens. Actuators 3, 14. R. Jha and A. K. Sharma, “High-performance sensor based
79–88 (1982). on surface plasmon resonance with chalcogenide prism and
2. K. Jindal, M. Tomar, R. S. Katiyar, and V. Gupta, “N-doped aluminum for detection in infrared,” Opt. Lett. 34, 749–751
ZnO thin film for development of magnetic field sensor based (2009).
on surface plasmon resonance,” Opt. Lett. 38, 3542–3545 (2013).
15. T. Srivastava, R. Jha, and R. Das, “High-performance bimetal-
3. K. Lin, Y. Lu, J. Chen, R. Zheng, P. Wang, and H. Ming,
lic SPR sensor based on periodic-multilayer-waveguides,”
“Surface plasmon resonance hydrogen sensor based on metal-
IEEE Photon. Technol. Lett. 23, 1448–1450 (2011).
lic grating with high sensitivity,” Opt. Express 16, 18599–
16. P. K. Maharana, S. Bharadwaj, and R. Jha, “Electric field
18604 (2008).
enhancement in surface plasmon resonance bimetallic con-
4. C. Perrotton, R. J. Westerwaal, N. Javahiraly, M. Slaman, H.
figuration based on chalcogenide prism,” J. Appl. Phys. 114,
Schreuders, B. Dam, and P. Meyrueis, “A reliable, sensitive
014304 (2013).
and fast optical fiber hydrogen sensor based on surface plas-
mon resonance,” Opt. Express 21, 382–390 (2013). 17. T. Srivastava, R. Das, and R. Jha, “On the high performance
5. G. Nemova, A. V. Kabashin, and R. Kashyap, “Surface of channel photonic crystal waveguide comprising different
plasmon-polariton Mach-Zehnder refractive index sensor,” J. plasmonic active metals,” Appl. Phys. B 108, 629–634 (2012).
Opt. Soc. Am. B 25, 1673–1677 (2008). 18. G. Hass, “On the preparation of hard oxide films with pre-
6. D. C. Cullen, R. G. Brown, and C. R. Lowe, “Detection of cisely controlled thickness on evaporated aluminum mirrors,”
immuno-complex formation via surface plasmon resonance on J. Opt. Soc. Am. 39, 532–540 (1949).
gold-coated diffraction gratings,” Biosensors 3, 211–225 (1987). 19. S. Roh, T. Chung, and B. Lee, “Overview of the characteristics
7. M. J. Jory, P. S. Vukusic, and J. R. Sambles, “Development of of micro- and nano-structured surface plasmon resonance
a prototype gas sensor using surface plasmon resonance on sensors,” Sensors 11, 1565–1588 (2011).
gratings,” Sens. Actuators B 17, 203–209 (1994). 20. P. K. Maharana and R. Jha, “Chalcogenide prism and
8. C. R. Lawrence, N. J. Geddes, D. N. Furlong, and J. R. Sambles, graphene multilayer based surface plasmon resonance affin-
“Surface plasmon resonance studies of immunoreactions uti- ity biosensor for high performance,” Sens. Actuators B 169,
lizing disposable diffraction gratings,” Biosens. Bioelectron. 161–166 (2012).
11, 389–400 (1996). 21. W. K. Jung, N. H. Kim, and K. M. Byun, “Numerical study on
9. D. Zhang, P. Wang, X. Jiao, G. Yuan, J. Zhang, C. Chen, H. an application of subwavelength dielectric gratings for high-
Ming, and R. Rao, “Investigation of the sensitivity of H-shaped sensitivity plasmonic detection,” Appl. Opt. 51, 4722–4728
nano-grating surface plasmon resonance biosensors using (2012).
rigorous coupled wave analysis,” Appl. Phys. A 89, 407–411 22. M. G. Moharam and T. K. Gaylord, “Rigorous coupled-wave
(2007). analysis of metallic surface-relief gratings,” J. Opt. Soc. Am.
10. K. H. Yoon, M. L. Shuler, and S. J. Kim, “Design optimization A 3, 1780–1787 (1986).
of nano-grating surface plasmon resonance sensors,” Opt. 23. S. Janz, Y. Beaulieu, A. Fiore, P. Bravetti, V. Berger, E.
Express 14, 4842–4849 (2006). Rosencher, and J. Nagle, “Surface emitted second-harmonic
11. K. M. Byun, S. J. Kim, and D. Kim, “Grating-coupled generation from a quasi-phase matched waveguide in
transmission-type surface plasmon resonance sensors based an AlxCa1-xAs/Al2O3 microcavity,” Opt. Express 2, 462–470
on dielectric and metallic gratings,” Appl. Opt. 46, 5703–5708 (1998).
(2007). 24. W. Zhu, C. J. Hirschmugl, A. D. Laine, B. Sinkovic, and S. S. P.
12. D. Cai, Y. Lu, K. Lin, P. Wang, and H. Ming, “Improving the Parkin, “Determination of the thickness of Al oxide films used
sensitivity of SPR sensors based on gratings by double-dips as barriers in magnetic tunneling junctions,” Appl. Phys. Lett.
method (DDM),” Opt. Express 16, 14597–14602 (2008). 78, 3103–3105 (2001).

1554 APPLIED OPTICS / Vol. 54, No. 6 / 20 February 2015