scieee Science in your language
[en] (orig)

Three-dimensional superresolution in metamaterial slab lenses: Experiment and theory

Abstract

This paper presents a theoretical and experimental study on the viability of obtaining two- and three-dimensional superresolution (i.e., resolution overcoming the diffraction limit for all directions in space) by means of metamaterial slab lenses. Although the source field cannot be actually reproduced at the back side of the lens with superresolution in all space directions, the matching capabilities of metamaterial slabs does make possible the detection of images with three-dimensional superresolution. This imaging takes place because of the coupling between the evanescent space harmonic components of the field generated at both the source and the detector

Read accessible full text

Three-dimensional superresolution in metamaterial slab lenses: Experiment and theory

Author: Mesa Ledesma, Francisco Luis; Freire Rosales, Manuel José; Marqués Sillero, Ricardo; Baena, J.D.
Publisher: American Physical Society
Year: 2005
DOI: 10.1103/PhysRevB.72.235117
Source: https://idus.us.es/bitstreams/01bc11bb-d0ea-4fd1-8cc8-956f4519f1b8/download
Th ee-dimensional supe esolu ion in me ama e ial slab lenses: Expe imen and heo y
F. Mesa
Depa men de Física Aplicada I. Uni e sidad de Se illa, 41012-Se illa, Spain
M. J. F ei e, R. Ma qués, and J. D. Baena
Depa men de Elec ónica y Elec omagne ismo, Uni e sidad de Se illa, 41012-Se illa, Spain
共Recei ed 29 Sep embe 2005; published 27 Decembe 2005兲
This pape p esen s a heo e ical and expe imen al s udy on he iabili y o ob aining wo- and h ee-
dimensional supe esolu ion 共i.e., esolu ion o e coming he di ac ion limi o all di ec ions in space兲by
means o me ama e ial slab lenses. Al hough he sou ce ield canno be ac ually ep oduced a he back side o
he lens wi h supe esolu ion in all space di ec ions, he ma ching capabili ies o me ama e ial slabs does make
possible he de ec ion o images wi h h ee-dimensional supe esolu ion. This imaging akes place because o
he coupling be ween he e anescen space ha monic componen s o he ield gene a ed a bo h he sou ce and
he de ec o .
DOI: 10.1103/PhysRe B.72.235117 PACS numbe 共s兲: 42.30.Wb, 41.20.Jb, 73.20.M , 78.20.Ci
I. INTRODUCTION
I is well known om he ea ly wo ks o Veselago1 ha a
slab made o a le -handed medium will ocus he elec o-
magne ic ene gy coming om a poin sou ce o ano he poin
loca ed a he opposi e side o he slab. An expe imen al
con i ma ion o his ocusing o ene gy has been epo ed by
Houck e al.2Subsequen wo ks3–5 ha e shown ha , unde
some ci cums ances, me ama e ial lenses made o le -
handed media can p oduce images a ce ain planes wi h a
esolu ion beyond he classical di ac ion limi , o “supe -
esolu ion imaging” 共SRI兲. This SRI has been a ibu ed o an
ampli ica ion, inside he lens, o he e anescen space Fou-
ie ha monics 共SFHs兲coming om he sou ce.3I has been
also discussed6 ha his p ocess gi es ise o ields ha decay
exponen ially om he lens owa d he image, which causes
supe esolu ion o ake place only in planes pa allel o he
slab in e aces. In he di ec ion pe pendicula o he lens, a
s ong decay o he ield is obse ed, and hus, a h ee-
dimensional 共3D兲pic u e o he sou ce canno be di ec ly
ob ained om he ield dis ibu ion a he back side o he
lens. In o he wo ds, supe esolu ion in he ans e se di ec-
ions is ob ained a he p ice o a d as ic loss o esolu ion in
he longi udinal di ec ion. This ac has been expe imen ally
co obo a ed by ecen ly epo ed ield measu emen s,4
whe e he ield g ow h om he image plane o he “supe
lens” can be clea ly app ecia ed. O he expe imen al esul s
also lead o he same conclusion, showing ha images “o
ini e dep h”5canno be di ec ly ob ained om ield measu e-
men s in SRI expe imen s.
Following he seminal pape o Pend y,3o he supe eso-
lu ion de ices ha e been p oposed making use o slabs o
nega i e pe mi i i y,7,8 e i e slabs,9coupled plana
pola i on- esonan me asu aces,10 and a pai o coupled
magne oinduc i e su aces.11 In spi e o he di e en physi-
cal na u e o hei cons i u i e elemen s, all hese s uc u es
p esen se e al common cha ac e is ics. In his way, he
abo e well-de ined amily o imaging de ices can be cha -
ac e ized by: 共i兲 he p esence o plana slabs, 共ii兲 he dis ance
be ween he sou ce and he image planes is always 2d共whe e
dis he slab wid h兲, and 共iii兲 he imaging p ope ies a e based
on he ampli ica ion o e anescen SFHs inside he slab.
Consequen ly, he imaging p ope ies o Pend y’s le -handed
slab lens a e also expec ed o be sha ed by hese de ices.6
Ne e heless, some expe imen al esul s ecen ly epo ed by
some o he au ho s11 ha e sugges ed he possibili y o ob-
aining supe esolu ion in he longi udinal di ec ion in one o
he de ices men ioned abo e. In hese expe imen s, a 3D
map o a poin like sou ce 共i.e., a sou ce o subwa eleng h
size兲was ob ained a he image side o he lens. In o he
wo ds, h ee-dimensional supe esolu ion imaging 共3D-SRI兲
was ob ained. The aim o he p esen con ibu ion is o p o-
ide he gene al heo y unde lying his 3D-SRI. I will be
shown ha 3D-SRI o poin like sou ces is ac ually a gene al
p ope y o he a o emen ioned amily o de ices, p o ided
ha he app op ia e de ec ion p ocedu e is ollowed.
II. ANALYSIS
The main di e ence be ween SRI and s anda d imaging
p ocesses is ha , in he o me , he in o ma ion o he image
o ma ion is ca ied ou by e anescen SFHs,3which can
suppo subwa eleng h in o ma ion, unlike he p opaga i e
SFHs esponsible o he con en ional imaging, which can
only anspo o e wa eleng h in o ma ion. Since e anescen
ields canno ca y powe and any image measu emen e-
qui es some powe ansmission, i is appa en ha he de-
ec ion p ocedu e has o somehow “c ea e” a a eling powe
lux. This combined e ec should hen pe u b, subs an ially,
he ields a ound he de ec o in a way simila o ha ound
in he unneling e ec . As is well known, he unneling o
powe is due o he exci a ion o a pai o e anescen elec-
omagne ic wa es, whose in e e ence gi es ise o a non a-
nishing lux o powe . “Pe ec unneling” o powe in a
wa eguide illed by a me ama e ial has been ecen ly e-
po ed by some o he au ho s.12 In his wo k i is shown ha
maximum unneling o powe in a se up wi h iden ical inpu
and ou pu wa eguides is achie ed when he ou pu is placed
a a dis ance om he inpu equal o ha om he sou ce o
PHYSICAL REVIEW B 72, 235117 共2005兲
1098-0121/2005/72共23兲/235117共6兲/$23.00 ©2005 The Ame ican Physical Socie y235117-1
he image in a me ama e ial supe lens. This sugges s ha a
simila e ec could ake place in a me ama e ial supe lens,
p o ided ha a de ec o iden ical o he sou ce is used o he
measu emen s. In he ollowing, how o ake ad an age o
such an e ec in o de o ob ain 3D-SRI wi h me ama e ial
slabs will be shown.
The p esen s udy s a s wi h he canonical p oblem o he
o ma ion o images by a le -handed slab o hickness d
cha ac e ized by
⑀
/
⑀
0=
␮
/
␮
0=−1+i
␦
, whe e
␦
Ⰶ1 accoun s
o he necessa y losses ac o o a oid he di e gence o
ield in eg als.13–15 In ou s udy, and ollowing a usual p o-
cedu e in mic owa e SRI expe imen s, he sou ce will be an
an enna 共speci ically, a loop an enna兲whose plane is loca ed
pa allel o he slab in e aces as shown in Fig. 1. The ield
beyond he lens will be scanned by an ou pu an enna ha
plays he ole o de ec o . The sou ce employed he e is
equi alen o an homogeneous su ace dis ibu ion o mag-
ne ic dipoles gi en by
Ms=
再
I0e−i
␻
,inside he loop
0, ou side,
冎
共1兲
whe e Ideno es he ampli ude o he imposed ime-ha monic
cu en in he loop. The compu a ion o he longi udinal mag-
ne ic ield Hza he image side o he slab 共z⬎⌬+d兲can be
ca ied ou by means o he ollowing double-in e se Fou ie
ans o m:
Hz共x,y,z兲=1
4
␲
2
冕冕
dkxdkyG
˜
共kx,ky;z兲M
˜
s共kx,ky兲ei共kxx+kyy兲,
共2兲
whe e G
˜
共kx,ky;z兲is he Fou ie ans o m o he G een’s
unc ion o he s uc u e unde s udy and M
˜
s共kx,ky;z兲, he
Fou ie ans o m o he spa ial su ace dis ibu ion o mag-
ne ic dipoles. A e applying he duali y p inciple o exp es-
sion 共6兲in Re . 15 and aking ⌬=d/2, he Fou ie ans o m
o he G een’s unc ion is ound o be
G
˜
共kx,ky;z兲=−共kx
2+ky
2兲2
␤
e−i
␤
0共z+d兲
␻
兵共
␤␮
0−
␤
0
␮
兲2ei
␤
d−共
␤␮
0+
␤
0
␮
兲2e−i
␤
d其,
共3兲
whe e
␤
=冑
␻
2␧
␮
−kx
2−ky
2
␤
0=冑
␻
2␧0
␮
0−kx
2−ky
2.
Since Ez=0 in he p esen case, all he emaining ield com-
ponen s can be deduced om Hz.15
The nume ical compu a ion o 共2兲p o ides he map de-
pic ed in Fig. 2 o he magni ude o Hzin he 共x−z兲plane a
he back side o he lens. I should be no ed ha , gi en ha
he ope a ion equency is 3 GHz 共␭0=100 mm兲and ha d
and ⌬a e aken much less han he ee-space wa eleng h,
he ay model app oxima ion canno be employed he e o
ob ain he ields in he conside ed egion 共in o he wo ds, we
a e dealing wi h nea ields and he e o e in a SRI si ua ion兲.
Acco ding o p e ious discussions, Fig. 2 shows ha he ield
magni ude has a s ong decay along he zdi ec ion, so ha
no in o ma ion abou he loca ion o he sou ce in he longi-
udinal di ec ion can be ex ac ed om he ield dis ibu ion.
A mo e de ailed pic u e o he ield dis ibu ion a wo planes
z=共2±
␦
兲dis plo ed in Fig. 3 共wi h
␦
se o 0.3兲and com-
pa ed o he ield ampli ude nea he sou ce a z=±
␦
d共 he
ield dis ibu ions a z=
␦
dand z=−
␦
da e iden ical兲.Asis
expec ed om he p ope ies o he me ama e ial slab, he
ield dis ibu ion a he z=共2+
␦
兲dplane is almos iden ical o
ha a he z=
␦
dplane, hus con i ming ha he z=
␦
dand he
z=共2+
␦
兲dplanes a e, in ac , conjuga e planes. On he con-
a y, he ield dis ibu ion a he z=共2−
␦
兲dplane subs an-
FIG. 1. Geome y o a me ama e ial lens. The sou ce is a loop
an enna loca ed a z=0. The de ec o is also an iden ical loop an-
enna. The lens is o med ei he by a le -handed me ama e ial slab
o by ano he plana de ice ha p oduces SRI h ough ampli ica ion
o e anescen ha monics. The ou pu an enna can e en ually be
loaded wi h mic owa e esis o s in o de o minimize ield
pe u ba ion.
FIG. 2. 共Colo online兲Mapo log
10兩Hz兩co esponding o he
ield gene a ed a he back side o he lens by a sou ce loop an enna
shown in Fig. 1. The slab is made o a le -handed medium o
hickness d=4 mm wi h
⑀
/
⑀
0=
␮
/
␮
0=−1+i0.001, and is sepa a ed
by a dis ance ⌬=d/2 om he sou ce. The sou ce is conside ed, o
simplici y, a ci cula loop an enna o adius 1=5 mm. The wi e
adius is o 0.2 mm. The ope a ion equency is 3 GHz.
MESA e al. PHYSICAL REVIEW B 72, 235117 共2005兲
235117-2
ially di e s in magni ude om he ield dis ibu ion a he
z=−
␦
dplane, nea he sou ce. These ac s co obo a e he
a o emen ioned discussions: he supe esolu ion in he ans-
e se di ec ions 共 he la e al dimensions o he sou ce and he
image a e app oxima ely one- en h o he wa eleng h兲is
compensa ed by an almos comple e loss o esolu ion in he
longi udinal di ec ion.
The p oblem o he measu emen o he image in he SRI
se up unde s udy is nex conside ed. In he mic owa e
ange, he image de ec ion is pe o med by measu ing he
ansmission coe icien be ween a sou ce an enna and a e-
cei ing an enna, which is scanned in he image side o he
lens.2,4,5 Fo simplici y, he ecei ing 共o ou pu 兲an enna is
assumed o be iden ical o he inpu an enna employed as
sou ce. The ansmission coe icien is measu ed by connec -
ing he inpu an enna o a wa e gene a o ia a wa eguide,
and he ou pu one o a de ec o h ough ano he iden ical
wa eguide 共mo e de ails o his measu emen se up a e e-
po ed in Re . 11兲. In his app oach, he me ama e ial slab
should be iewed as a ma ching de ice,16 whose ansmis-
sion coe icien 共o S21 in he usual mic owa e e minology兲
is gi en by17
=2Z12Z0
共Z11 +Z0兲共Z22 +Z0兲−Z12
2,共4兲
whe e Zij a e he elemen s o he impedance ma ix o he
sys em o med by he wo an ennas and he le -handed slab,
and Z0is he cha ac e is ic impedance o he inpu and ou pu
wa eguide. In o de o measu e a supe esolu ion image, he
size o he loop an ennas has o be smalle han he ee-
space wa eleng h, which addi ionally would make he eal
pa o Zij 共 he adia ion esis ance17兲be negligible wi h e-
spec o i s imagina y pa ; namely, Zij⯝−i
␻
Lij, whe e Lij is
he induc ance ma ix o he sys em. The diagonal e ms o
he induc ance ma ix co espond o he induc ances o a
single-loop an enna aced o he le -handed slab. Howe e ,
in he “pe ec lens” con igu a ion conside ed he e, he slab
does no a ec he ields a ound he sou ce3and, he e o e,
L11=L22 ⬅L, whe e Lis he sel -induc ance o he loops in
ee space. The nondiagonal e ms o he induc ance ma ix
accoun o he mu ual induc ance o he loop an ennas in he
p esence o he slab, namely, L12 =L21⬅M. Thus, he ans-
mission coe icien in 共4兲can be w i en as
=2i
␻
MZ0
␻
2共L2−M2兲+2i
␻
LZ0−Z0
2.共5兲
Since Ldoes no change wi h he posi ion o he an ennas,
he spa ial dependence in 共5兲comes only om M. The de-
pendence o 兩 兩wi h he eac ances X11=X22 =−
␻
Land X12
=X21=−
␻
Mdeduced om 共5兲is shown in Fig. 4. I can be
seen how he maximum ansmission 共兩 兩⬇1兲is achie ed
when X11⯝X12, excep o e y small alues o he a io
X11/Z0. This e ec is illus a ed in Fig. 5, whe e he alues o
X12 and X11, which co esponds o a maximum o he ans-
mission coe icien , a e plo ed. The e o e, excep when
␻
L
ⰆZ0, he ansmission coe icien eaches i s maximum a
hose poin s whe e M⬇L共兩 兩⬇1i LⲏZ0兲. Since bo h he
inpu and ou pu an ennas a e iden ical, and he ield a he
sou ce plane is ep oduced a he image plane 共see Fig. 3兲,
FIG. 3. Plo o he log10兩Hz兩a di e en z-planes o Fig. 1. The
slab and he sou ce loop an enna a e as in Fig. 2.
FIG. 4. Plo o he modulus o he ansmission coe icien
=S21 as a unc ion o he modulus o he eac ances X11=X22=
−
␻
Land X21=−
␻
M.
FIG. 5. Plo o he alues o he eac ances X11=X22=−
␻
Land
X112=X21=−
␻
M o which he modulus o he ansmission coe -
icien 共5兲 eaches a maximum.
THREE-DIMENSIONAL SUPERRESOLUTION IN…PHYSICAL REVIEW B 72, 235117 共2005兲
235117-3
he condi ion M⬇Lis expec ed o be sa is ied a ound he
poin 共x,y,z兲=共0,0,2d兲. In consequence, a maximum o he
ansmi ed powe should be de ec ed a ound his pa icula
poin , and hus, i would appea as an “e ec i e ocusing
poin ” o he pe ec lens.
In o de o show he abo e expec ed e ec s in a p ac ical
si ua ion, he magni ude o he ansmission coe icien has
been nume ically compu ed o he con igu a ion unde
s udy. The impedance ma ix has been calcula ed by impos-
ing known cu en s, Ii=1,0 A, in he loops, and hen com-
pu ing he co esponding sel and mu ual induc ances. The
compu a ion o he sel -induc ance is an s anda d elec o-
magne ic p oblem, and he mu ual induc ance is ob ained
a e compu ing he lux o he magne ic ield gi en by 共2兲
ac oss he su ace o he ou pu an enna. The ansmission
coe icien is inally de e mined om 共4兲, assuming Z0
=50 ⍀. In Fig. 6, a map o he compu ed ansmission coe -
icien o a sys em composed o wo iden ical lossless loop
an ennas is shown. I can be obse ed ha his igu e sub-
s an ially di e s om Fig. 2; in pa icula , a clea maximum
o 兩 兩can be obse ed in Fig. 6 in he neighbo hood o he
image a 共0,0,2d兲. These esul s clea ly show he di e ence
be ween he ansmi ed powe and he ield dis ibu ion in
he absence o he ou pu an enna, and also how 3D-SRI can
be ob ained when he app op ia e de ec o is used.
Le us now conside he measu emen o he unpe u bed
ield. Fo his pu pose, he de ec o should be designed o
a ec he ield dis ibu ion as li le as possible. I could be
closely achie ed by loading he ou pu an enna wi h an ad-
di ional high esis ance, which will signi ican ly educe he
cu en induced a he ou pu an enna, as well as i s gene a ed
ield. In Fig. 7 he compu ed alues o he ansmission co-
e icien along he z-axis o Fig. 1 a e shown o di e en
esis ances, R, loading he ou pu an enna. As is expec ed, he
cu e o R=0 ⍀shows a maximum nea he loca ion o he
image, which hus appea s as a “ ocusing poin ” o he lens,
a z⬇2d.共The small di e ence be ween he ac ual loca ion
o his maximum wi h espec o z=2dcan be a ibu ed o
he no e y high alue o he a io
␻
L/Z0⬇8 in he case
unde s udy兲. As is also expec ed, he cu es o he highes
alues o R esemble he beha io o 兩Hz兩in Fig. 2 o he
unpe u bed con igu a ion.
Ano he ob ious s a egy o educing he pe u ba ion o
he ield by he measu emen is o educe he adius o he
ou pu an enna o Fig. 1. This will educe bo h L2共 he induc-
ance o he ou pu an enna兲and Mwi hou changing L1共 he
induc ance o he inpu loop兲. In he limi when
␻
L2,
␻
M
ⰆZ0,共4兲 educes o
⬇2i
␻
M
i
␻
L1−Z0
.共6兲
Since bo h L1and Z0do no depend on he loca ion o he
ou pu , he measu ed ansmission coe icien u ns ou o be
p opo ional o M, i.e. o he magne ic ield Hza he loca ion
o he ou pu . Thus, he analyzed expe imen al se up is ap-
p op ia e o he de ec ion o he ield o he unpe u bed
sys em o med by he inpu an enna and he lens. The mag-
ni ude o he ansmission coe icien along he zaxis o Fig.
1 is plo ed in Fig. 8 o se e al alues o he adius o he
ou pu loop. The cu es co esponding o 2艌4 mm clea ly
show ha he loca ion o he de ec ed ansmission maximum
app oaches he lens in e ace as he adius o he ou pu loop
is educed. In he limi o small adius 共 2=1 mm兲, he p e-
iously p edic ed mono onic inc ease o he ield owa d he
igh -hand-side lens in e ace is obse ed. Figu e 8 also
shows ha he sensi i i y o he expe imen al se up o small
di e ences be ween he shapes o he inpu and ou pu an-
ennas is no e y high. As i can be eadily seen om he
cu es co esponding o 2=5 and 4.5 mm, he ansmission
peak is loca ed e y close o 2d o mode a e de ia ions o 2
om i s op imum alue 2= 1.
III. EXPERIMENT
In o de o show he applica ion o he abo e heo y o
p ac ical de ices, he expe imen al demons a ion o 3D-SRI
ecen ly epo ed by some o he au ho s11 will now be eex-
amined in ligh o he abo e conside a ions. Al hough he
FIG. 6. 共Colo online兲Map o he magni ude in decibels o he
ansmission coe icien , 兩S21兩, be ween he inpu and ou pu an en-
nas o Fig. 1. The s uc u al pa ame e s a e as in Fig. 2. The ou pu
an enna is iden ical o he inpu one.
FIG. 7. Magni ude o he compu ed ansmission coe icien
along he Zaxis be ween he inpu and ou pu an ennas o Fig. 1 o
di e en esis ances loading he ou pu an enna. The s uc u al pa-
ame e s a e as in Fig. 2.
MESA e al. PHYSICAL REVIEW B 72, 235117 共2005兲
235117-4
unde lying physics o he de ice analyzed he e is no ex-
ac ly he same as ha o he le -handed pe ec lens,11 bo h
sys ems show he same p ocess o ampli ica ion o e anes-
cen modes and a e equi alen in p ac ice o he p esen
pu poses. Thus, expe imen al esul s plo ed in Fig. 9 show
he loca ion o he maximum o he ansmission coe icien
along he zaxis o he magne oinduc i e lens11 o di e en
alues o he esis ance loading he ou pu loop. The expe i-
men al se up used o ob ain hese esul s is desc ibed in de-
ail in Re . 11. The only di e ence is ha he ou pu an enna
is now loaded by di e en mic owa e esis o s o ob ain he
esul s shown in Fig. 9. These esul s show he same ype o
beha io as ha epo ed in Fig. 7; namely, he loca ion o a
maximum o he magni ude o he ansmission coe icien
in he neighbo hood o z⬇2d o iden ical ou pu and inpu
an ennas, and a simila displacemen o he maximum o he
ansmission coe icien as he adius o he ou pu an enna
dec eases. The ag eemen be ween he heo y and he expe i-
men s can be conside ed as a alida ion o he p esen heo y.
I also shows ha he ideas de eloped in he p e ious analy-
sis a e e y gene al, and ha , ac ually, hey a e applicable o
any supe esolu ion lens based on he ampli ica ion and am-
pli ude es o a ion o FHs along he de ice.
IV. CONCLUSIONS
Supe esolu ion in me ama e ial supe lenses has been in-
es iga ed. As is well known, his imaging is p ima ily due o
he ampli ica ion o e anescen modes inside he lens. As a
consequence, supe esolu ion in he planes pa allel o he
slab is una oidably compensa ed by a d as ic loss o esolu-
ion in he di ec ion pe pendicula o he slab. Howe e ,
since e anescen modes do no ca y powe , any physical
de ec ion 共namely, a measu emen 兲o he image a he back
side o he lens will equi e he exis ence o some amoun o
ansmi ed powe om he sou ce o he de ec o , which can
signi ican ly a ec he o iginal ield dis ibu ion. I his las
e ec is aken in o accoun , he seconda y ields gene a ed by
he p esence o he de ec o should be conside ed. Following
his app oach, i has been shown ha he ansmission coe -
icien be ween he sou ce and he de ec o can be inc eased
conside ably. P o ided ha he app op ia e de ec o is em-
ployed 共 ypically, a lossless ou pu an enna iden ical o he
inpu an enna兲, his maximum will occu in a neighbo hood
o he image, hus esul ing in supe esolu ion imaging also
in he di ec ion pe pendicula o he slab. Howe e , i should
be emphasized ha o ob aining his 3D-SRI, some p e ious
knowledge o he sou ce is necessa y in o de o design he
de ec o p ope ly. Thus, a gene al conclusion a ises om he
analysis: Supe esolu ion in me ama e ial supe lenses is al-
ways incomple e; i he dis ance om he sou ce o he lens is
known, he shape and cha ac e is ics o he sou ce can be
eco e ed wi hou unce ain y om he analysis o he ield
dis ibu ion a he image side o he lens. Con e sely, i he
shape and cha ac e is ics o he sou ce a e known, i is pos-
sible o design an app op ia e de ec o in o de o localize
he sou ce spa ially om he alues o he ansmission co-
e icien be ween he sou ce and he de ec o . Howe e , i is
impossible o de e mine, simul aneously, by means o a
me ama e ial supe lens, he loca ion, shape, and cha ac e -
is ics o an unknown sou ce wi h a esolu ion o e coming he
di ac ion limi . We eel ha he epo ed analysis and ex-
pe imen s, as well as he conclusions a ising om hem, will
be o impo ance in he design o me ama e ial supe esolu-
ion de ices.
ACKNOWLEDGMENT
This wo k has been suppo ed by he Spanish Minis y o
Educa ion and Science by p ojec Con ac No. TEC2004-
04249-C02-02.
FIG. 8. Magni ude o he compu ed ansmission coe icien
along he zaxis be ween he inpu and ou pu an ennas o Fig. 1 o
se e al alues o he adius o he de ec o 共 2兲. The s uc u al pa-
ame e s a e as in Fig. 2.
FIG. 9. Plo s o he measu ed magni ude o he ansmission
coe icien be ween he inpu and ou pu an ennas o Fig. 1 when
he slab is subs i u ed by he magne oinduc i e lens epo ed in Re .
11. 共F equency=3 GHz兲.
THREE-DIMENSIONAL SUPERRESOLUTION IN…PHYSICAL REVIEW B 72, 235117 共2005兲
235117-5

1V. G. Veselago, So . Phys. Usp. 10, 509 共1968兲.
2A. A. Houck, J. B. B ock, and I. L. Chuang, Phys. Re . Le . 90,
137401 共2003兲.
3J. B. Pend y, Phys. Re . Le . 85, 3966 共2000兲.
4A. G bic and G. V. Ele he iades, Phys. Re . Le . 92, 117403
共2004兲.
5A. N. Laga ko and V. N. Kissel, Phys. Re . Le . 92, 077401
共2004兲.
6N. Fang and X. Zhang, Appl. Phys. Le . 82, 161 共2003兲.
7N. Fang, H. Lee, C. Sung, and X. Zhang, Science 308, 534
共2005兲.
8R. J. Blaikie and D. O. S. Mel ille, J. Op . A, Pu e Appl. Op . 7,
S176 共2005兲.
9R. Ma qués, F. Mesa, and F. Medina, Appl. Phys. Le . 86,
023505 共2005兲.
10 S. Maslowski, S. T e yako , and P. Ali alo, J. Appl. Phys. 96, 161
共2004兲.
11 M. F ei e and R. Ma qués, Appl. Phys. Le . 86, 182505 共2005兲.
12 J. D. Baena, L. Jelinek, R. Ma qués, and F. Medina, Phys. Re . B
72, 075116 共2005兲.
13 N. Ga cía and M. Nie o-Vespe inas, Phys. Re . Le . 88, 207403
共2002兲.
14 D. R. Smi h, D. Schu ig, M. Rosenblu h, S. Schul z, S. Anan ha
Ramak ishna, and J. B. Pend y, Appl. Phys. Le . 82, 1506
共2003兲.
15 R. Ma qués and J. D. Baena, Mic owa e Op . Technol. Le . 41,
290 共2004兲.
16 V. G. Veselago, cond-ma /0501438 共unpublished兲.
17 D. M. Poza , Mic owa e Enginee ing 共Wiley, New Yo k, 1998兲,
2nd ed.
MESA e al. PHYSICAL REVIEW B 72, 235117 共2005兲
235117-6