Surface plasmon resonance from metallic sculptured thin films

A. Shalabaney,1 I. Abdulhalim,1* A. Lakhtakia,2,3 A. Lahav,1 Christian Patzig,4 I. Hazek,1 A. Karabchevky,1 Bernd Rauschenbach,4 F. Zhang,2 J. Xu 2

1 Department of Electrooptic Engineering, Ben Gurion University of the Negev, Beer Sheva

84105, Israel 2 Department of Engineering Science and Mechanics, Pennsylvania State University, University

Park, PA 16802-6812, USA 3 Department of Physics, Indian Institute of Technology Kanpur, Kanpur 208016, India

4 Leibniz-Institut für Oberflächenmodifizierung e.V., Permoserstrasse 15, 04318 Leipzig, Germany

*Corresponding author: abdulhlm@bgu.ac.il

[Ibrahim: All changes are in red. All comments are in purple. Akhlesh]

Abstract: Surface plasmon (SP) waves can exist on the interface of a dielectric (such as water)

and a metallic sculptured thin film (STF) of porosity as high as 0.55. As the porosity increases,

the SP resonance (SPR) dip widens, shifts to higher wave numbers, and becomes asymmetric due

to increasing scattering losses. Above 0.6 porosity, the SPR dip disappears, leaving behind only

a peak near the onset to the total-internal-reflection regime. The shape of the nanoislands

constituting the STF is better described as ellipsoidal than as spherical or spheroidal, indicating

thereby the existence of orientational biaxial anisotropy even in such very thin films. For a best

fit between the theoretical calculations and the experimental data the STF is divided into two

layers having different porosity and nanoisland shape. Sensitivity of the STF-based SPR signal

to refractive index variations of an analyte infiltrating the nanopores of and in the region

adjoining the metal STF is found to be doubly enhanced compared to that for the SPR signal

from a homogeneous metal film.

1. Introduction

Metallic nanostructured thin films are of immense interest these days1 due to their widely

increasing technological importance for applications such as solar-control mirrors,

subwavelength optical imaging, and biosensing. This wide interest can be attributed to surface

plasmon (SP) waves. The phenomenon of surface plasmon resonance (SPR) has been known for

a long time2 and is commonly used for sensing of chemicals3. In order to excite an SP wave, the

tangential component of the wave vector of the p-polarized incident light has to match the SP

wave vector in magnitude. This match can be achieved using either a high-index prism at angles

larger than the critical angle for total internal reflection (TIR) or a one-dimensional grating. A

thin metal film is usually deposited on the prism or the grating and then p-polarized reflected

light exhibits a sharp dip at the angle of incidence at which the component of the incident light’s

wave vector along the surface closely matches the wave number of the SP wave. This resonance

occurs due to the real part of the relative permittivity of the metal being negative at optical

frequencies.

A typical set-up for SPR sensors is the Kretschmann configuration, wherein one side of a

high-index prism is coated with a thin metal film, typically silver or gold. The metal plays two

roles: First, excitation through the prism sets up evanescent fields that extend to the metal-

sample interface; second, at that interface, SP waves can be excited when the tangential

component of the wave vector of the incident light matches the SP wave vector in magnitude.

The reflectance strongly decreases when the latter condition (called the SPR condition) is met

due to absorption in the metal, and the width and the depth of the resonance are dependent on the

type of the adjacent dielectric material (analyte) which could be air. This sensing method is

sensitive to surface defects (due to oxidation of the metal film) and surface roughness, since the

field strength is largest near the metal-analyte interface and the penetration depth in the analyte is

rather small. Nonetheless, the width of the reflectance dip and the level of the reflectance

minimum do yield information about the analyte, and the accuracy of the SPR sensor can be

enhanced by optimizing the thickness of the metal film4.

Figure 1: (a) Schematic of the oblique-angle-deposition technique. Collimated vapor flux

oriented at angle with respect to the substrate plane leads to the formation of parallel

columns tilted at an angle to the same plane. (b)-(f) Typical scanning-electron-

microscope images of used columnar thin films (CTFs) made of (b) aluminum, (c) silver (cross

section), (d) silver (top view), (e) gold (cross section), and (f) gold (top view). The scale bar in

all of the micrographs corresponds to 100nm.

Figure 2: Schematic of the Kretschmann configuration. The drawing is not to scale. The higher

the value of the angle for SPR, the higher is the SP wave number.

Surface plasmons can be localized on metallic nanoparticles and nanoshells5,6. Localized

SPR (LSPR) can also arise at some locations either in or on certain porous or highly disordered

materials7. The associated absorption resonances often show up in optical spectra as narrowband

features. Spatial nonhomogeneities can also lead to averages over many different localized

resonances, the overall effect being broadened thereby.

We demonstrate here the existence of SPR on the surface of a nanoengineered sculptured thin

film (STF). Sufficiently thick STFs are assemblies of upright, parallel, shaped nanowires

generally grown by physical vapor deposition (PVD) techniques8. The nanoscale morphology of

STFs comprises clusters of 3-5 nm diameter. STFs can be made of inorganic and organic

dielectric materials, polymers, metals, and semiconductors. For optical purposes, an STF is a

unidirectionally nonhomogeneous continuum with anisotropic constitutive properties at visible

and infrared wavelengths.

STFs can be nanoengineered in a variety of morphologies and made of a variety of materials.

A hallmark of an STF is its porosity with 10-to-300-nm void regions between nanowires of

similar cross-sectional dimensions. The porosity can be engineered to lie within the 0.1-to-0.9

range. Being porous, an STF can function as a nano-reactor; that capability can be harnessed for

a variety of sensing applications. Very thin STFs (< 50 nm) may be considered as assemblies of

tilted nanoislands.

Recently, LSPR was excited in porous metallic thin films––without prism coupling––due to

scattering9,10 and surface-enhanced phenomena such as Raman scattering11,12,13 and

fluorescence14 were demonstrated. Extensive SPR experiments using prism coupling were

performed in the past on films comprising metallic nanoislands15,16 and long-range SPR17

excitation was also demonstrated; however, those films were quite dense with porosity less than

0.2. In the majority of these experiments the theoretical fit to the experimental data was

performed using the Maxwell Garnett [Ibrahim: There is no hyphen between the two names

of James Clerk Maxwell Garnett. Akhlesh] effective-medium formalism which works best

when the porosity is either very low or very high. Otherwise, it better to use the Bruggeman

formalism set up correctly for an anisotropic effective medium18,19.

We demonstrate here the excitation of SPR on the surfaces of porous metallic STFs, using

prism coupling in the visible regime. We also found that absorption in nonporous metal films

can produce a sharp and symmetric peak as the signature of the SPR phenomenon. Hence the

effect of two mechanisms of loss on the SPR was examined in thin metal films––absorption and

scattering. Both mechanisms were found to cause broadening of the SPR dip remaining with a

peak at the onset of TIR.

2. Experimental details

The STFs chosen for this purpose are called columnar thin films (CTFs), and were made of

aluminum, silver, or gold. The CTFs made of aluminum were deposited in an electron-beam

evaporation system (PVD-75, KJL Inc.) at the Pennsylvania State University. The oblique-angle-

deposition technique was used, as depicted schematically in Fig. 1(a). With the vacuum base-

pressure set below 4 µTorr in an evacuated chamber, collimated aluminum vapor was directed

towards a 2.54-cm × 2.54-cm BK7-glass substrate at a fixed angle 20o to the substrate plane.

The distance between the aluminum source and the centroid of the substrate was fixed at 25.4

cm. The substrate was held stationary. The deposition rate, monitored with a resonating quartz

crystal sensor, was controlled at 0.25 nm/s. After deposition, a profilometer (Tencor P-10) was

used to measure the metal CTF’s average thickness as 30 nm.

CTFs of gold and silver were grown at the Leibnitz Institute; the gold CTFs were grown by

dc sputtering, and silver CTFs by electron-beam evaporation20,21,22. The vapor was directed very

obliquely (

�

χv as low as 5°) to the substrate plane. Table 1 summarizes the samples used and

their parameters. [Ibrahim: I think that the second column of Table 1 contains 90° –

�

χ1 and

�

χ1; likewise the third column is also confused. Please be careful. My program uses

�

χ1 and

�

χ2. These angles are measured from the substrate plane (not for the substrate normal).

Note that the vapor incidence angles are

�

χv1 and

�

χv2. These angles do not appear in my

program. Also note that

�

χ1 ≥ χv1 and

�

χ2 ≥ χv2. So Table 1 needs to be redone carefully.

Akhlesh] In Figs. 1(b)-(f) we present typical scanning electron microscope (SEM) micrographs

showing the tilted columnar morphology of all the CTFs we used. For reference measurements,

dense films of each material were also prepared with vapor directed normally (

�

χv = 90°)

towards the substrate plane.

[Ibrahim: Columns 8 and 11 in Table 1 are redundant, and should be thrown out.

Akhlesh]

[Ibrahim: In Figure 10, “Por1” should be “p1”, “Por 2” should be “p2”, “Reflectivity”

should simply be “R”. What is “na” in Fig. 10?

The Kretschmann configuration of Fig. 2 was implemented with incident light coming from a

653-nm-wavelength laser diode first collimated with a lens and then p-polarized after passage

through a linear polarizer. The 45o-90o-45o prism used in the Kretschmann configuration was

made of BK7 glass (refractive index = 1.51509). The glass substrate on which the metal CTF

was deposited was separated from the prism by a thin layer of an index–matching fluid of index

1.52. The fraction R of the incident power density that exits the right slanted face of the prism

was measured using an amplifying photodiode and an oscilloscope. The entire assembly was

mounted on an optical table on which the angle of incidence on the prism can be set to 1o-

precision. To ensure that the SPR phenomenon was correctly observed, we rotated the polarizer

by 90o degrees to the TE polarization and verified that the SPR does not then exist.

Figure 3: Measured values of R versus (denoted by E), when the metal film is (a) a pre-etching

aluminum CTF and (b) a post-etching aluminum CTF. The dashed lines are best-fit numerical

simulations obtained with (a) p = 0.517 and (b) p = 0.560. The sample to be sensed is air.

3. Description of the numerical simulations

Numerical simulations of R as a function of were performed as follows: The p–polarization

transmittance Tag across the air-glass interface on the left side of the prism was computed as a

function of with the Fresnel formula, as also the p–polarization transmittance Tga at the glass-

air interface on the right side of the prism. The p-polarization reflectance Rgca of glass-CTF-

sample at the base of the prism was computed as a function of the incidence angle , by using a

simple matrix-based formalism wherein we assumed that the nanowires of the CTF are tilted in

the plane of Fig. 2 at an angle , with respect to the substrate plane [7, Chap. 8]. The

reflectance R was estimated as the product TagRgcaTga, while the porosity p of the CTF and the

angle of incidence were varied to fit the experimental data.

The effective relative permittivity tensor of the aluminum CTF was obtained from a

Bruggeman formalism18,19, 23 , wherein it was assumed that the metal is distributed in the form of

electrically small ellipsoids of aspect ratios

�

1:γ 2 :γ 3, the void regions are electrically small

spheres, the void regions are vacuous, the porosity p lies between 0 and 1. It should be

mentioned that Yang et al.17 used the Maxwell Garnett approach to perform their theoretical fit

to experimental data. Furthermore, although Yang et al.10 used the Bruggeman approach for the

same purpose, they assumed that both the metal and air (void) are distributed as identical,

electrically small, prolate spheroids because they took the effective medium to be uniaxial. Thus

our Bruggeman approach is substantially different as it allows the effective medium to be

biaxial. Another point is that in reference 10 they assumed the trace of the depolarization dyadic

to be unity which was shown by Lakhtakia et. al19 to be not true for spheroids and ellipsoids.

[Ibrahim: I checked with paper and pen, finally. Their depolarization dyadic and ours

differ by a factor such that the previous sentence is incorrect. Of course, they used prolate

spheroids, whereas we are more general. Akhlesh] Moreover, while Yang et al. 10 treated the

depolarization dyadic (of a prolate spheroid) as an unknown to which they fit their experimental

data, we assume that both shape factors and

�

γ 3 of ellipsoids are unknowns.

4. Results and discussion

For aluminum CTFs, the relative permittivity of bulk aluminum equals −59.288 + 22.2385i at

653-nm wavelength, and we set the tilt angle , film thickness as 30 nm, and the

ellipsoids of aspect ratios 1:1.2:15. Figure 3(a) shows the measured values of R versus the

incidence angle , when the analyte is air. The sudden drop in R as changes from 5o to 0o

indicates the excitation of an SPR. The wide dip in the curve is indicative of scattering

loss due to the spatial nonhomogeneity of matter in the CTF. Fitting the numerically simulated

reflectance R as a function of the internal angle of incidence to the experimental data, we

found that p = 0.517; i.e., the CTF is 51.7% volumetrically porous.

Next, we etched the aluminum CTF using the Transene24 AL etchant of type A, for 5 s. The

etching rate at 25 oC was 1 nm/s. The CTF was then rinsed in water and dried by blowing dry

nitrogen on it. Thereafter, R was measured again in the setup over the same range of . The

measured data is presented in Fig. 3(b). Clearly, the SPR dip is very different from that in the

Fig. 3(a). Numerical simulation of R and fitting to the experimental data suggests that p = 0.56.

In other words, etching increased the porosity from 0.517 by 0.043, and the effect was captured

by a widening of the SPR dip.

CTFs prepared by the oblique-angle-deposition technique are highly discontinuous when

their thickness is less than 50 nm, as they are made of nanoislands but with some orientational

order which builds up more and more as nanowires as the thickness increases25. As SPR in the

Kretschmann configuration had been previously excited only with very dense thin films, the

possibility of its excitation using metallic CTFs less than 50 nm thick was questionable. Here we

have now shown both theoretically and experimentally that SPR excitation is possible on sub-50-

nm thick aluminum CTFs with porosity ~0.5. As the porosity increases, the SPR dip widens and

becomes asymmetric due to increasing scattering losses in the CTF due to the nonhomogeneous

distribution of matter therein. Numerical simulations (not shown) indicate that as the porosity

increases beyond 0.75, the SPR dip almost disappears, with a vestigial peak near the onset to the

TIR regime, as also seen in Figure 3 [Ibrahim: The last clause is not correct, because Fig. 3

does not have any vestigial peak. In order to make that claim, a figure like the solid line in

Fig. 5 of our recent Proc. SPIE paper is needed. Akhlesh].

Figure 4: Measured and simulated data for p-polarization reflectance R versus the angle , when

the nonporous metal film (p = 0) is 8 nm thick and made of chromium. The curves exhibit a

broad dip and a sharp and symmetric peak due to the absorption in bulk chromium. The analyte

sensed is either air or water (refractive index = 1.33).

Next, numerical simulations were carried out with the assumptions that (i) the analyte is

water of refractive index 1.33, and (ii) the void regions of the aluminum CTF are completely

filled with the analyte. Figures 3(c) and (d) show the results for p = 0.517 and 0.56, the same

values as obtained for Figs. 3(a) and (b), respectively. A comparison of the two sets of figures

reveals that the SPR peak shifts by ~37.5o when the refractive index of the same changes by

0.33, thereby indicating a strong sensitivity of ~113 deg/RIU, larger than the sensitivity of the

SPR from a dense aluminum film (~79 deg/RIU) [Ibrahim: The previous number needs a

reference to support it. Akhlesh]

The appearance of a peak at the onset of the TIR regime when the SPR broadens can also be

seen when the broadening is due to absorption loss rather than scattering loss inside the thin

metal film. Aluminum and chromium, for examples, have bulk refractive indexes with high

imaginary parts at red wavelengths (3.102 + 3.334i for Cr and 1.39 + 7.65i for Al); therefore,

absorption loss is strong in these metals. In order to verify whether the broadening of the SPR

dip can also be due to absorption loss, we both carried out experiments and performed

simulations on SPR from thin films of nonporous (p = 0) chromium and aluminum. There

cannot be scattering loss in a film with p = 0.

Figure 4 shows curves for an 8-nm-thick chromium film deposited on a microscope

slide and incorporated in the Kretschmann configuration of Fig. 2. The dark blue and pink

curves are simulations for the sample being air and water, respectively. For these simulations, we

used the 2x2 Abeles matrix method.26 For the experiment with water, a plastic tank was filled

with water and glued to the microscope slide with silicone rubber. The other two full curves in

Fig. 4 connect the experimentally obtained data points. As the angle

is varied, a strong peak appears followed by a deep dip in each

Figure 5: Intensity distribution of the reflected diverging beam for the two polarization states: (a)

p and (b) s. A peak area is clearly seen for the p polarization state. [Ibrahim: TM and TE have

not been used in this paper. Please replace by p and s in the figure above. Akhlesh]

curve. Comparison between the results for the two analytes (air and water) indicates a strong

sensitivity of the angle to the refractive-index change (~93 deg/RIU) in the region set aside for

occupation by an analyte. Figure 5 shows the intensity distribution of a diverging beam showing

a peak region in the TM case which does not appear in the TE polarization case.

Since the CTFs were made of aluminum, we also performed simulations for SPR from

nonporous films of aluminum, as shown in Fig. 6. Due to absorption, a sharp and symmetric

peak is obtained for nonporous aluminum films as thin as 4 nm; for thicker films, the SPR dip

appears. A sharp and symmetric peak is obtained with a thickness of 30 nm for the CTF with

~0.3 porosity, while with the nonporous film only a 4 nm thickness is required. This observation

should be correlated with absorption loss in a CTF (p ~ 0.3) being much lower than

Figure 6: Numerical simulations of the p-polarization reflectance R versus the angle , when the

metal film is nonporous aluminum of different thicknesses. Broad SPR dips exist for all

thicknesses, and a sharp and symmetric peak is obtained for 4-nm thickness. The analyte is air.

in a nonporous film (p = 0) of the same metal. But porosity also induces SPR broadening due to

scattering loss inside a porous film. How can we distinguish between the roles of porosity and

intrinsic absorption in the appearance of the SPR phenomenon? In order to understand the issue,

we simulated the p-polarized reflectance from nonporous gold films, because bulk gold has an

imaginary part of the refractive index almost half as that of aluminum (0.1498 + 3.423i). As can

be seen in Fig. 7, a sharp and symmetric peak is obtained with 25-nm-thick nonporous gold film

which is not very different in thickness from a 30-nm-thick, 30% porous, aluminum CTF. We

cannot deduce from this fact that the losses in the nonporous gold film and in the aluminum CTF

are similar because the reflection level around the remnant peak is different. Comparing now

Figs. 6 and 7, we notice that the higher the loss, the lower the reflectance R on both sides of the

peak. This is due to the fact that both at the onset of TIR and in the excitation of the SPR, we

have a surface wave which gets absorbed or scattered more, depending on the nature of the loss

on the metallic side of the metal-sample interface. If so, then we can conclude from Figs. 3 and

6 that the aluminum CTF is more dissipative than the nonporous aluminum film.

Figure 7: Same as in Fig. 6 for nonporous gold film. Broad SPR dips exist for all thicknesses,

and a sharp and symmetric peak is obtained for 25-nm thickness. The analyte is air.

[Ibrahim: I am unable to see Figs. 6 and 7 in the document that you sent me. Akhlesh]

Assuming an exponentially decaying wave, we get from the reflectance near the sharp and

symmetric peaks for aluminum films in Figs. 3 and 6 that:

, , (1)

where is the attenuation coefficient––which is basically an absorption coefficient for the

nonporous film, but includes a scattering coefficient for the CTF, i.e.,

––while denotes the distance traveled by the

surface wave along the metal-analyte interface. Solving these equations, we obtain the following

optical densities: and . Assuming the absorption

part is similar for the porous and nonporous films (i.e., ),

we get which is about 43% of the absorption part and is due to the

porosity of the CTF.

Table 1: Parameters used for best fit between the theoretical calculations and the experimental

data presented in Figs. 3 and 8. [Ibrahim: In the names of samples, please replace “STF” by

“CTF”. A similar change in figures is also warranted. Akhlesh]

Silver and gold possess an extinction coefficient smaller by about factor of 2 than that of

aluminum at the wavelength of interest. Hence SPR dip broadening is expected to be primarily

induced by the disordered nanostructure. As can be seen in Fig. 7 it is possible to get a

[Ibrahim: I have to stop here. I cannot see many figures completely. Please take care of

the issues that I have raised above as well as in the list of references, and send me the next

version. In the figures below, please use “phi” instead of “internal angle” and “R” instead

of “R-TM”. As much as possible, these figures should look like Fig. 3; or Fig. 3 should look

like these figures. At that time, please send me a pdf too. Thanks. Akhlesh Jan 12, 2009]

Figure 8: TM reflectivity using the SPR setup of Fig. 2 from silver films. Dots are experimental

data points while curves are theoretical fits using the parameters listed in Table 1.

symmetric peak at the TIR onset when the closed gold film thickness is 25nm, however its height

relative to the reflection background is 45% [Ibrahim: Since R lies between 0 and 1, please

replace 45% by 0.45, etc., for consistency. Akhlesh] and the SPR dip is still prominent. This

is due to the relatively smaller extinction coefficient in comparison to the closed aluminum case

shown in Fig. 6 which shows a TIR peak height of 80% and broad SPR dip for aluminum

thickness of 4nm. Several silver and gold STF samples were prepared at the IoM for this

purpose as listed in table 1. Figures 8 and 9 show the reflectance curves obtained using the same

setup of Fig. 2 where the solid curves are the theoretical fits obtained using the parameters listed

in table 1. In order to get the best fits the layer was considered as composed of two STF layers

having different porosities

Figure 9: TM reflectivity using the SPR setup of figure 2 from the gold films. Dots are

experimental data points while curves are theoretical fits using the parameters listed in table 1.

The dashed curves in (a) and (d) correspond to different parameters (2nd set in table 1).

and ellipsoidal rods with different aspect ratios. Note that the first layer which is in the range 3-

15nm, has higher porosity than the 2nd layer. A closer look at the cross section images in fig. 1c

and 1e indicates a more porous region near the substrate indeed. For samples Ag-STF-3a, Au-

STF-4a and Au-STF-3a the SPR dip becomes broad and less pronounced, however it clearly

exists. The theoretical fits show that high porosities exist in these films with the possibility of

unknown inhomogeneity in Au-STF-4a and Au-STF-3a samples (figure 9) as the fits are not as

Figure 10: Simulation results to demonstrate the sensitivity enhancement as the porosity

increases using silver double-layer STF with similar parameters as of sample Ag-STF-2a. (a)

TM reflectivity vs. internal angle when the two layers have similar porosities. (b) Calculated

sensitivity when the porosity of the 1st layer (from the prism side) was kept constant as indicated

in the legend while the porosity of the 2nd layer is varied from 0 till 0.3.

good as with the cases of low porosity samples. For samples Au-STF-3a and Au-STF-5a the

best fit was found by using a set of film parameters that are different by more than 30% from the

nominal parameters particularly in film thickness and porosity.

One of our motivations for using STFs in sensing is their porous nature. The porosity

increases the surface to volume ratio and thus the sensitivity is expected to increase. This was

demonstrated in figure 3 for the case of Al-STF when the SPR dip turns into peak at the TIR

onset. To investigate the sensitivity enhancement as a function of the porosity we considered

the structure of the sample Ag-STF-2a. TM reflectivity curves for different porosities of the STF

film are shown in figure 10a for two indices (1.325 and 1.332) of the water sample. Note that the

reflectivity drops drastically for angles above 85 degrees due to the fact that very little energy

passes through the prism at these large angles because in these graphs the product TagRgcaTga is

plotted. In figure 10b the sensitivity is plotted versus the porosity of the 2nd layer for different

fixed values of the porosity of the 1st layer. Note that as the porosity increases the sensitivity

increases and enhanced by about factor of 2 for porosities around 30%. Similar results obtained

using Au STFs. Hence using STFs made of gold or silver one can gain enhancement of the

sensitivity of SPR sensors. The fact that the dip becomes wider as the porosity increases can be

overcame by using subpixelling algorithms as was demonstrated recently27.

5. Concluding remarks

The excitation of SPR from porous columnar thin films of metals was shown to be possible,

using the Kretschmann configuration. The SPR dip widens as the porosity increases until it

disappears, leaving behind only a peak near the onset of the TIR regime. The excitation of the

SP wave at the onset of the TIR regime, combined with scattering and/or absorption losses

causes the reflectance for p-polarization to decrease for angles of incidence on the right side of

the peak while the SPR excitation causes it to decrease on the left side of the peak. Under certain

conditions, one gets the two sides of the peak to be equally steep and the peak is then symmetric

in shape. Experimental data indicates that, due to absorption loss, a nonporous thin film of

chromium that is just 8 nm thick produces a symmetric peak in air as well as in water, with high

sensitivity. Hence, this approach of using lossy metals with optimum thickness to obtain a

narrow symmetric peak widens the possibilities for SPR sensors by making available a larger

variety of materials. Although nonporous lossy metals are known for long time, they have not

been used as SPR sensors because of their wide SPR dip. Here we proposed their use as sensors

but based on the remnant peak at the TIR onset. The use of porous CTFs has the advantage of

allowing larger-sensitivity sensors as the ratio of surface to volume increases; in the case of 30-

nm porous aluminum films, the sensitivity using the SPR peak mode is enhanced by a factor of

about 1.5 in comparison to standard SPR sensors. For silver and gold films the SPR dip

sensitivity increase by about factor of 2 with 30% porosity as compared to closed metal.

Acknowledgments

The work at BGU is supported by the Israeli Ministry of Science under the “Tashtiot” program.

A. Lakhtakia thanks the Charles Godfrey Binder Endowment at the Pennsylvania State

University for partial support.

References:

1 S. Lal, S. Link, and N. J. Halas, “Nano-optics from sensing to waveguiding,” Nature Photon. 1,

641-648 (2007).

2 H. Raether, Surface Plasmons on Smooth and Rough Surfaces and on Gratings (Springer,

Berlin, 1988).

3 I. Abdulhalim, M. Zourob, and A. Lakhtakia, “Surface plasmon resonance sensors: A mini-

review,” Electromagnetics 28, 214-242 (2008).

4 P. Lecaruyer, M. Canva, and J. Rolland, “Metallic film optimization in a surface plasmon

resonance biosensor by the extended Rouard method,” Appl. Opt. 46, 2361-2369 (2007).

5 K. L. Kelly, E. Coronado, L. Zhao, and G. C. Schatz, “The optical properties of metal

nanoparticles: The influence of size, shape, and dielectric environment,” J. Phys. Chem. B 107,

668-677 (2003).

6 S. A. Kalele, N. R. Tiwari, S. W. Gosavi, and S. K. Kulkarni, “Plasmon-assisted photonics at

the nanoscale,” J. Nanophoton. 1, 012501 (2007).

7 A. K. Sarychev and V. M. Shalaev, “Field distribution, Anderson localization, and optical

phenomena in random metal-dielectric films,” In: Optics of Nanostructured Materials, edited

by V. A. Markel and T. George (Wiley, New York, NY, USA, 2000) pp. 227-282.

8 A. Lakhtakia and R. Messier, Sculptured Thin Films: Nanoengineered Morphology and Optics

(SPIE Press, Bellingham, WA, USA, 2005).

9 A. I. Maaroof, A. Gentle, G. B. Smith and M. B. Cortie, “Bulk and surface plasmons in highly

nanoporous gold films,” J. Phys. D: Appl. Phys. 40, 5675-5682 (2007).

10 F. Yang, G. W. Bradberry, and J. R. Sambles, "The study of the optical properties of obliquely

evaporated nickel films using IR surface plasmons, Thin Solid Films, 196, 35-46 (1991).

11 M. Suzuki, W. Maekita, Y. Wada, K. Nakajima, K. Kimura, T. Fukuoka, and Y. Mori, “In-line

aligned and bottom-up Ag nanorods for surface-enhanced Raman spectroscopy,” Appl. Phys.

Lett. 88, 203121 (2006).

12 S. Shanmukh, L. Jones, J. Driskell, Y. Zhao, R. Dluhy, and R. A. Tripp, “Rapid and sensitive

detection of respiratory virus molecular signatures using a silver nanorod array SERS

substrate,” Nano Lett. 6, 2630-2636 (2006).

13 S. Ahl, P. J. Cameron, J. Liu, W. Knoll, J. Erlebacher, and F. Yu, "A comparative plasmonic

study of nanoporous and evaporated gold films," Plasmonics 3, 13-20 (2008).

14 I. Abdulhalim, A. Karabchevsky, C. Patzig, B. Rauschenbach, and B. Fuhrmann,

“Comparative study of enhanced fluorescence from nano sculptured thin films,” Proc. SPIE,

7041, 70410G (2008).

15 T. Inagaki, J. P. Goudonnet, P. Royer, and E. T. Arakawa, “Optical properties of silver island

films in the attenuated-total-reflection geometry,” Appl. Opt. 25, 3635-3639 (1986).

16 J. W. Little, T. A. Callcott, T. L. Ferrell, and E. T. Arakawa, "Surface-plasmon radiation from

ellipsoidal silver spheroids," Phys. Rev. B 29, 1606 (1984).

17 F. Yang, G. W. Bradberry and J. R. Sambles, "Long-range surface mode supported by very

thin silver films," Phys. Rev. Lett. 66, 2030-2032 (1991).

18 A. Lakhtakia, B. Michel, and W. S. Weiglhofer, "The role of anisotropy in the Maxwell

Garnett and Bruggeman formalisms for uniaxial particulate composite media," J. Phys. D:

Appl. Phys. 30, 230-240 (1997).

19 W.S. Weiglhofer, A. Lakhtakia, and B. Michel, “Maxwell Garnett and Bruggeman formalisms

for a particulate composite with bianisotropic host medium,” Microw. Opt. Technol. Lett. 15,

263-266 (1997); correction: 22, 221 (1999).

20 E. Schubert, T. Höche, F. Frost, and B. Rauschenbach, “Nanostructure fabrication by glancing

angle ion beam assisted deposition of silicon,” Appl. Phys. A 81, 481-486 (2005).

21 Ch. Patzig, B. Rauschenbach, W. Erfurth, and A. Milenin, “Ordered silicon nanostructures by

ion beam induced glancing angle deposition,” J. Vac. Sci. Technol. B 35, 833-838 (2007).

22 B. Fuhrmann, H. S. Leipner, H. R. Höche, L. Scubert, P. Werner, and U. Gösele, “Ordered

arrays of silicon nanowires produced by nanosphere lithography and molecular beam epitaxy,“

Nano. Lett. 5, 2524-2527 (2005).

23 J. A. Sherwin, A. Lakhtakia, and I. J. Hodgkinson, “On calibration of a nominal structure–

property relationship model for chiral sculptured thin films by axial transmittance

measurements,” Opt. Commun. 209, 369-375 (2002).

24 www.transene.com/aluminum.html

25 R. Messier and J. E. Yehoda, “Geometry of thin-film morphology,” J. Appl. Phys. 58, 3739-

3746 (1985).

26 F. Abéles, “ ………. , ” Ann. Phys. (Paris) 5, 596-640 (1950). [Ibrahim : Please complete.

Akhlesh]

27 A. Lahav, M. Auslender, and I. Abdulhalim, "Sensitivity enhancement of guided wave surface

plasmon resonance sensors," Opt. Lett. 33, 2539-2541 (2008).