1 Surface Plasmon-Polariton Waves
(a) Excitation of SPP waves:
Figure 1: Dispersion relationship
Generating SPP requires energy and momentum conservation and this is done via coupling based on total internal reflection. So, as indicated below:
Figure 2: Evanescent waves between two media
θ > θc= arcsinn2 n1
, we have TIR.
In mediumn1:
ω
cn1=q
k2k+k2z kk = ωn1
c sinθ In mediumn2:
kk is conserved:
kk= ωn1
c sinθ
A propagating wave inn2 would have
ω
cn2 =kk
ω
cn1sinθ >
ω
cn2=⇒n1sinθ > n2
θ >arcsinn2
n1 =θc
Because the wave is evanescent, we must go close to the SPP wave.
Figure 3: Excitation mechanisms of SPP We need to create an evanescent wave in the dielectric.
Figure 4: Dispersion relationship for SPP excitation
βo= ωo
c nprismsinθ=⇒θ
This means for each EM wave at frequencyω, there is an angle at which the energy and in-plane momentum are conserved, hence, the coupling occurs, as indicated in Fig. 5.
The reflectance drops at angleθ because of the energy transfer from EM wave to the surface plasmon.
(b) Propagation length of a SPP wave.
(ω) = 1− ω2p
ω(ω+iγ) =R+iI
Figure 5: Reflectance as a function of angle For a complex refractive index:
β+i kI = ω
c(nR+i nI ), wherekI is related tonI
ei(β+ikI)z =e−kIzeiβz I(z) =Ioe−2kIz(Intensity)
Figure 6: Propagation length as a function of propagation axis z We assume that losses are small: γ ω, ωp
βR+βI ' ω c
r md
m+d
If we notice from that equation thatβ has now also a complex value becausem has complex value. d=1 for simplicity
βR+βI = ω c
s
(R+iI).1 R+iI+ 1
= ω c
s R+iI
R+ 1(1 +iI
R+1)
=βR
v u u u u t
1 +iI
R
1 +i I R+ 1
Now we can use approximation, e.g. we can considerI <|R| βR+βI 'βR
r
(1 +iI R
)(1−i I R+ 1)
∼βR r
1 +iI( 1
R − 1 R+ 1)
∼βR
1 +iI
2 ( 1
R − 1 R+ 1)
βI ' I
2βR 1 R(R+ 1) l= 1
2βI = R(R+ 1) IβR We replaceβR with
ω c
r R
R+ 1 l= R(R+ 1)
I
c ω
rR+ 1 R
l'(c ω)1
I
√R(R+ 1)32
(ω) = 1− ω2p
ω(ω+iγ) =⇒R= 1− ωp2 ω2+γ2 andI = ωp2
ω(ω2+γ2)γ l= c
ω ω3 ωp2
s 1− ω2p
ω2(2− ω2p ω2)32γ−1
l=c ω
ωp
2s 1−ωp2
ω2 2−ω2p ω2
!3
2
γ−1
1.1 Evanescent Wave
Figure 7: Evanescent waves
Under total internal reflection, the wavevector in medium n2 reads:
ω2
c2n22 =k2k+k2z k2k = ω2
c2n21sin2θ
=kz2ω2
c2(n22−n21sin2θ)
kz= ωn2
c s
1−n21sin2θ n22 But:sin2θ >
n2
n1 2
=⇒1−n21
n22 sin2θ <1 kz =iωn2
c s
n21sin2θ n22 −1 E(x, z) =Eoeikkxeikzz
E(x, z) =Eo e
−ωcn2
s
n2 1 n2 2
sin2θ−1
eikkx
The enclosed part of the above equation is what gives evanescent wave. The figure below shows the behavior of evanescent waves.
Figure 8: Exponential decay of Evanescent waves
2 Scattering and absorption cross sections of a metal nanoparticle
Here we have a nanoparticle, with radius ’a’ being hit by an EM plane waves as depicted below.
The generated polarizabilityp=oαEo, whereα= αo
1−ik3
6παo
and αo= 4πa3m−1
m+2
Incident Field: Ei=Eo
Scattered Field (from the dipole): Es= 4πP
o(...) Total Field: Et=Ei+Es
Figure 9: EM plane waves impinged onto a nanoparticle Power through the integration sphere reads:
P = Z
4π
S~tot.·~nr2dΩ S~tot = 1
2Re{Etot×Htot∗ } S~tot =S~inc+S~sca+S~int
where
S~inc= 1
2Re{Einc×Hinc∗ }, S~sca= 1
2Re{Esca×Hsca∗ }, S~int= 1
2Re{Esca×Hint∗ }+1
2Re{Eint×Hsca∗ } Pinc=
Z
4π
S~inc·~nr2dΩ = 0.
This is because the EM field coming from outside to the nanoparticle goes out. In other words, no field remains in the nanoparticle.
Psca = Z
4π
S~sca·~nr2dΩ =scattered power Pint=
Z
4π
S~int·~nr2dΩ6= 0
P =Psca+Pint=−d
dtEsphere ,
whereEsphereis the energy in the integration sphere. The variation of the energy in the volume is related to the power absorbed in the sphere, so−dtdEsphere =Pabs where Pabs is the power absorbed by the particle−Pabs =Psca+Pint with rearrangement of the terms, we get
−Pint=Psca+Pabs Psca+Pabs =Ptot ,
this is the total power dissipated by the induced dipole. If we consider the nanoparticle as a dipole, then
Ptot=ωImn
~ p ~Ei∗o
=ωoIm nE~iE~i∗
o
=ωo |E~i|2 Im{α}
Incident intensity: Ii = Z1 |E~i|2, whereZ =qµ
o
o and it is the vacuum impedance Ptot =ωoZIiIm{α}
Ptot=IikIm{α}
σtot =kIm{α}
cross-section, which has the dimensions of an area, and it expresses the effective area of the nanoparticle when it interacts with light. Psca=Power scattered by~p= Iiσsca , whereasσsca
is the scattering cross-section.
This result can be obtained by calculated the power radiated by the dipole.
σsca= k4 6π |α|2 Pabs=Ptot−Psca= (kIm{α} − k4
6π |α|2)Ii
σabs is the absorption cross-section=kIm{α} − k4 6π |αo |2 Im{α}=Im
( αo
1−i6πk3αo )
= Im{αo}
1 +(6π)k62 |αo |2 + k3 6π
|αo|2 1 +(6π)k62 |αo|2 where
σtot =kIm{α} and σsca = k4 6π |α|2 σabs = kIm{αo}
1 +(6π)k62 |αo|2
σtot=σsca+σabs
−Pint=Psca+Pabs
Ptot=−Pint
From the figure above, the EM plane waves propagate toward the detector, therefore if we
Figure 10: Measurement of radiated power by the nanoparticle
integrate across the whole integration sphere; the only non-zero part will be the part at the detector. So, without nanoparticle the detector would measure the incident power. But with nanoparticle the detector would measurePinc+Pint+
Pscabecause the detector is small P =Pinc+Pint=Pinc−Psca−Pabs
P =Pinc−Ptot
If we make an experiment and measure the curve will look like the one shown below.
Figure 11: Exctinction of power at the detector position
The dip is called Extinction, whereas Ptot = Pext =⇒ The Extinction cross-section σext = σsca+σabs. From that we see in a simple experiment, we can measure the power dissipated by a nanoparticle through this extinction dip, and this is also called optical theorem. That theorem states that the dip measured in transmission by an infinitesimal detector is equal to the total power dissipated (scattering + absorption) by the nanoparticle.