Fabrication of Single-Electron Transistors Using Field

Delivered by Ingenta toJun-ichi SHIRAKASHI

IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE

Copyright copy 2010 American Scientific PublishersAll rights reservedPrinted in the United States of America

Journal ofNanoscience and Nanotechnology

Vol 10 7239ndash7243 2010

Fabrication of Single-Electron Transistors UsingField-Emission-Induced Electromigration

Watari Kume Yusuke Tomoda Michinobu Hanada and Jun-ichi Shirakashilowast

Department of Electrical and Electronic Engineering Tokyo University of Agriculture andTechnology Koganei Tokyo 184-8588 Japan

We report a novel technique for the fabrication of planar-type Ni-based single-electron transistors(SETs) using electromigration method induced by field emission current The method is so-calledldquoactivationrdquo and is demonstrated using arrow-shaped Ni nanogap electrodes with initial gap sep-arations of 21ndash68 nm Using the activation method we are easily able to obtain the SETs byFowler-Nordheim (F-N) field emission current passing through the nanogap electrodes The F-Nfield emission current plays an important role in triggering the migration of Ni atoms The nanogapis narrowed because of the transfer of Ni atoms from source to drain electrode In the activationprocedure we defined the magnitude of a preset current Is and monitored the current I betweenthe nanogap electrodes by applying voltage V When the current I reached a preset current Is westopped the voltage V As a result the tunnel resistance of the nanogaps was decreased from theorder of 100 T to 100 k with increasing the preset current Is from 1 nA to 150 A Especially thedevices formed by the activation with the preset current from 100 nA to 15 A exhibited Coulombblockade phenomena at room temperature Coulomb blockade voltage of the devices was clearlymodulated by the gate voltage quasi-periodically resulting in the formation of multiple tunnel junc-tions of the SETs at room temperature By increasing the preset current from 100 nA to 15 A inthe activation scheme the charging energy of the SETs at room temperature was decreased rang-ing from 1030 meV to 320 meV Therefore it is found that the charging energy and the number ofislands of the SETs are controllable by the preset current during the activation These results clearlyimply that the activation procedure allows us to easily and simply fabricate planar-type Ni-basedSETs operating at room temperature

Keywords Electromigration Field Emission Current Nanogap Single-Electron Transistor

1 INTRODUCTION

Fabrication and control of nanometer-scale devices suchas single-electron transistors (SETs) have recently becomepromising subjects of intensive study Room temperatureoperation of the devices is considered to be an importantmilestone in SETs history So far a number of experi-ments have been reported for the production of SETs usingnanofabrication scheme which involves either highly elab-orate electron-beam lithography1 or special techniquesand processes such as shadow evaporation2 deposi-tion of metal nanoparticles3ndash7 electromigration-inducedbreaking of thin metal wires defined by electron-beamlithography89 and resistive microstrips with no intentionaltunnel junctions1011 However the SETs fabricated bythese methods have been mainly restricted to the low tem-perature operation because of its small charging energy

lowastAuthor to whom correspondence should be addressed

Therefore in order to investigate planar-type SETs sophis-ticated nanofabrication scheme is required because of sub-nanometer dimension of the devices with higher operationtemperaturePreviously we have reported a wide-range control of

tunnel resistance of Ni nanogaps having less than 70 nmseparations using electromigration method induced byfield emission current12ndash14 By applying a voltage to theNi initial nanogaps Fowler-Nordheim (F-N) field emis-sion current flows through the nanogaps The Ni atomsat the tip of the source electrode are sufficiently acti-vated by the F-N field emission current for the migra-tion Afterwards the transfer of the Ni atoms fromsource to drain electrode is easily generated There-fore the separation of the nanogaps after performing themigration becomes narrower than that before the migra-tion resulting in a decrease of the tunnel resistance ofthe nanogaps We call this electromigration procedure

J Nanosci Nanotechnol 2010 Vol 10 No 11 1533-48802010107239005 doi101166jnn20102803 7239


IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE

Fabrication of Single-Electron Transistors Using Field-Emission-Induced Electromigration Kume et al

ldquoactivationrdquo Similar processes have been reported previ-ously in studies of the formation of Au bridges across70 nm slits15 and insulated planar nanoscopic contacts16

These phenomena can be understood on the basis of theelectrical breakdown procedure in planar contacts In thispaper we describe a fabrication scheme of planar-typeNiVacuumNi-based SETs with multiple tunnel junctionsusing activation method Currentndashvoltage (IndashV charac-teristics of the devices formed by the activation clearlyexhibited the suppression of electrical current at low-biasvoltages known as the Coulomb Blockade and Coulombblockade voltage was also modulated by the gate voltagequasi-periodically at room temperature Furthermore weobserved Coulomb staircase at 16 K The control proce-dure of single-electron charging energy and the number ofislands in the devices by the preset current of the activationis studied in detail

2 EXPERIMENTAL DETAILS

Activation experiments were performed on 100 nm thickthermally grown SiO2 surface on Si substrates First25 nm thick Au contact pads were fabricated upon 5 nmTi layer by electron-beam lithography and electron-beamevaporation Then 20 nm thick arrow-shaped Ni nanogapswith separations of 21ndash68 nm were formed by the sameprocedure in the gap between the Au contact pads Thearrow-shaped source-drain electrodes cause the concentra-tion of electric field at the tip of the nanogap electrodesFigure 1(a) shows a scanning electron microscope (SEM)image of the initial nanogap Before performing the acti-vation the nanogap in the device exhibited the separationof 33 nm with high tunnel resistance above 100 TFollowing these fabrication processes we applied the

activation procedure to the nanogaps Initial nanogapswere installed on a stage with manual probes in a vac-uum chamber with a pressure of 10minus3minus10minus4 Pa and wereoptionally selected from among the samples for the acti-vation procedure Through the experiments the activationprocedure was carried out in the vacuum chamber at roomtemperature The electrical properties of the nanogapsduring the activation were controlled and measured by a

Source

Drain

SiO2

Ni

33 nm

50 nm

Ni ClusterDot

lt10 nm

(a) (b)

Source

Drain

Ni

SiO2 50 nm

Fig 1 SEM image of nanogap electrodes (a) before and (b) after acti-vation with the preset current Is = 150 A

semiconductor parameter analyzer The detailed activationprocess for the fabrication of the SETs consists of fourmain steps(1) we set the magnitude of preset current Is(2) the voltage V is applied to the initial nanogap and isramped up while monitoring the current I passing throughthe nanogap(3) we stop the voltage V when the current I reaches apreset current Is and(4) after performing the activation drain current Id -drainvoltage Vd properties of the device are measured with themodulation of gate voltage Vg to confirm SET charac-teristics This procedure was continuously repeated to thedevices with increasing the preset current from 1 nA to150 A

3 RESULTS AND DISCUSSION

The activation procedure was performed for the fabricationof planar-type Ni-based SETs In the fabrication step ofthe SETs as mentioned above the voltage V was appliedto the nanogaps with the preset current Is from 1 nA to150 A Figure 1(a) shows a SEM image of Ni initialnanogap before the activation In the Figure initial gapseparation is estimated to be approximately 33 nm A SEMimage of the nanogap after the activation with the pre-set current Is = 150 A is exhibited in Figure 1(b) Theimage clearly indicates that the gap separation was nar-rowed from 33 nm to less than 10 nm by the accumulationof Ni atoms within the gap suggesting the formation of Niclusterdot structures between source and drain electrodes

102

104

106

108

1010

1012

1014

1016

10ndash10 10ndash9 10ndash8 10ndash7 10ndash6 10ndash5 10ndash4 10ndash3

T = 300 KNi 20 nm

Tunneling Metallic

A

B C

Tun

nel r

esis

tanc

e R

(Ω

)

Preset current Is (A)

Insulating

Fig 2 Dependence of tunnel resistances of nanogap electrodes on pre-set current Is The low-voltage tunnel resistance R is measured afterthe activation The separation of the initial gap is about 33 nm LabelsA B and C correspond to the devices activated with the preset currentIs = 400 nA 650 nA and 15 A respectively

7240 J Nanosci Nanotechnol 10 7239ndash7243 2010


IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE

Kume et al Fabrication of Single-Electron Transistors Using Field-Emission-Induced Electromigration

Figure 2 represents the relation between the tunnel resis-tance R of the nanogaps and the preset current Is duringthe activation procedure The tunnel resistance R whichis defined as the resistance of the nanogaps in the low-voltage regime was measured after performing the activa-tion The result suggests that the tunnel resistance R wasdecreased from 100 T to 100 k by increasing the presetcurrent Is from 1 nA to 150 A and was well controlledby the preset current Is Moreover a marked decrease inthe tunnel resistance R is clearly seen on the current rangeof 100 nA lt Is lt 10 A We defined the current rangeas ldquoTunnelingrdquo in which the electrical properties of thenanogaps vary between ldquoInsulatingrdquo and ldquoMetallicrdquo char-acteristics at room temperature12ndash14 From the point of viewof the fabrication scheme of SETs activation conditionswere investigated in ldquoTunnelingrdquo regime as shown withA B and C in Figure 2 First we studied the electricalcharacteristics of the nanogap activated with the preset cur-rent Is = 400 nA (ldquoArdquo in Fig 2) Figure 3(a) shows theIdminus Vd characteristics of the nanogap labeled ldquoArdquo as afunction of Vg and a current map measured as a func-tion of Vd and Vg (inset) In the device the gate voltageVg was swept from minus37 V to minus35 V by 01 V step atroom temperature In Figure 3(a) Coulomb blockade canbe observed in the IdminusVd characteristics correspondingto the charging energy Ec= 580 meV Moreover Coulombblockade voltage was also modulated by the gate voltageVg quasi-periodically resulting in the formation of mul-tiple tunnel junctions of the SETs In general Coulombblockade is clearly visible when the charging energy isabout ten to twenty times higher than the value of ther-mal fluctuation17ndash20 Hence the device can be operated asthe SETs at room temperature The same measurementprocedure was applied to the device after performing theactivation with the preset current Is = 650 nA as shownwith ldquoBrdquo in Figure 2 IdminusVd properties and a chargingdiagram were shown in Figure 3(b) For the increase ofthe preset current Is from 400 nA to 650 nA the tun-nel resistance of the device was reduced from 75 T to490 G as shown in Figure 2 Furthermore the charg-ing energy Ec of the device was also decreased from580 meV to 470 meV with increasing the preset currentIn spite of the decrease of the charging energy severalCoulomb peaks were obtained at room temperature asshown in the inset of Figure 3(b) This proves unambigu-ously that the device consists of metallic clustersdots con-nected to source and drain electrodes through the multipletunnel junctions rather than a single junction Figure 3(c)shows the electrical characteristics of the nanogap acti-vated with the preset current Is = 15 A (ldquoCrdquo in Fig 2)which is close to the preset current for ldquoMetallicrdquo activa-tion regime In the activation condition with Is = 15 Aelectrical properties of the device gradually approach toldquoMetallicrdquo properties Therefore the charging energy Ecof Figure 3(c) was further reduced to 400 meV Multiple

Coulomb diamonds were however still exhibited with thegate voltage Hence it is suggested that the Ni atoms tendto accumulate as Ni clustersdots in the nanogap betweensource and drain electrodes and act as multiple islands ofthe SETs Furthermore these results imply that the numberof islands can be controlled by the magnitude of the presetcurrent passing through the nanogap during the activationFigure 4 represents the fabrication conditions of the

SETs as a function of initial gap separation W of thenanogaps and the preset current Is of the activation

ndash300

ndash200

ndash100

0

100

200

300

05

Dra

in c

urre

nt I d

(fA

)

T = 300 K Is = 400 nANi 20 nm

Vg = ndash37~ndash35 V

∆Vg = 01 V step

Sample A(a)

Drain voltage Vd (V)ndash1

ndash05

0

05

1

ndash35ndash50

ndash25

0

25

50

Dra

in v

olta

ge V

d (V

)

Gate voltage Vg (V)

Dra

in c

urre

nt I d

(fA

)

ndash37 ndash365 ndash36 ndash355

ndash1 ndash05 0

ndash3

ndash2

ndash1

0

1

2

3

Drain voltage Vd (V)

T = 300 KNi 20 nmIs = 650 nA

Sample B(b)

Dra

in c

urre

nt I d

(pA

)

Vg = ndash40~40 V

∆Vg = 5 V step

ndash05

ndash025

0

025

05

ndash2

ndash1

0

1

2

Gate voltage Vg (V)

Dra

in c

urre

nt I d

(pA

)

Dra

in v

olta

ge V

d (V

)

ndash05 ndash025 0 02540200ndash20ndash40

ndash150

ndash100

ndash50

0

50

100

150


T = 300 KNi 20 nmIs = 15 microA

Sample C(c)

Dra

in c

urre

nt I d

(pA

)

Vg = ndash40~40 V∆Vg = 5 V step

ndash05

ndash025

0

025

05

ndash30

ndash15

0

15

30

Dra

in v

olta

ge V

d (V

)

Gate voltage Vg (V)

Dra

in c

urre

nt I d

(pA

)


Fig 3 Currentndashvoltage characteristics of samples (a) A (b) B and (c)C at room temperature for different gate voltages For clarity each curveis shifted 20 fA (sample A) 200 fA (sample B) and 10 pA (sample C)The insets show current maps of the drain current Id versus Vd andVg (stability diagram) The samples A B and C correspond to those inFigure 2

J Nanosci Nanotechnol 10 7239ndash7243 2010 7241


IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE


procedure Filled circles indicate that the devices showCoulomb blockade voltage in the IdminusVd characteristicsat room temperature Filled triangles and squares repre-sent the devices with non-linear Idminus Vd characteristicswithout Coulomb blockade properties and completely lin-ear IdminusVd characteristics at room temperature respec-tively The labels A B and C in Figure 4 correspond tothose in Figures 2 and 3 From the Figure the formationof the SETs operating at room temperature is achievedusing the preset current from 10minus7 A to 10minus5 A in com-bination with the initial gap separation of the nanogapsfrom 21 nm to 68 nm Particularly the devices labeled and didnrsquot show Coulomb blockade properties atroom temperature However when we cooled the devicesto 16 K the IdminusVd curves exhibited a clear suppressionof the conductivity around zero bias regime as shown inFigures 5(a b and c) The charging energy estimated fromthe IdminusVd curves of sample (Fig 5(a)) and sample (Fig 5(b)) was 400 meV and 300 meV at 16 K respec-tively This is due to the fact that the activation condi-tion of sample is closer to ldquoMetallicrdquo regime than thatof sample Furthermore Coulomb staircases and peri-odic differential conductance peaks can be seen at 16 Kin sample activated with the preset current Is = 70 Aas shown in Figure 5(c) This result suggests that theSET having asymmetrical two tunnel junctions with singleisland structure is formed using the activation conditionwithin ldquoMetallicrdquo regime It is noted that the chargingenergy of sample is determined to be 190 meV at 16 Kbut completely linear IdminusVd characteristics are obtainedat room temperature These results indicate that the charg-ing energy of the devices was decreased from 400 meV

10ndash8 10ndash7 10ndash6 10ndash5 10ndash4 10ndash3

Initi

al g

ap s

epar

atio

n W

(nm

)

LinearSET Non-linear

10

20

30

40

50

60

70

80


T = 300 KNi 20 nm

α

β

γ

A

B

C

Fig 4 Fabrication conditions of SETs as a function of initial gap sep-aration W and preset current Is The symbols indicate devices with SETproperties (bull) non-linear IndashV without SET properties ( and linear IndashVproperties ( Labels A B and C and and correspond to thosein Figures 2 and 3 and Figure 5 respectively

ndash15

ndash1

ndash05

0

05

1

15

050

50

100

150

200

Dra

in c

urre

nt I d

(pA

)


Con

duct

ance

dId

V (

pS)

T = 16 K

Ni 20 nm

Is = 1 microA

(a)

EC = 400 meV

ndash5

ndash25

0

25

5

0

1

2

3

Dra

in c

urre

nt I d

(pA

)

Con

duct

ance

dId

V (

nS)


T = 16 K

Ni 20 nm

Is = 3 microA

EC = 300 meV

(b)

ndash15

ndash10

ndash5

0

5

10

15

0

50

100

150

200

250

Dra

in c

urre

nt I d

(nA

)

Con

duct

ance

dId

V (

nS)


T = 16 K

Ni 20 nm

Is = 70 microA

EC = 190 meV

(c)

Sample β

Sample γ

ndash05 ndash025 0 025

05ndash05 ndash025 0 025


Sample α

Fig 5 Currentndashvoltage characteristics and corresponding differentialconductance of SET devices at 16 K Samples (a) (b) and (c) rep-resent devices activated with different preset currents Is = 1 A 3 Aand 70 A respectively



IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE


102

103

104

10ndash7 10ndash6 10ndash5 10ndash4


T = 300 K W = 21 nm W = 26 nm W = 33 nm

T = 16 K

W = 35 nm(Sample α)

W = 43 nm(Sample β)

W = 56 nm(Sample γ)

α β γ

AB

C

Cha

rgin

g en

ergy

Ec

(meV

)

Fig 6 Charging energy Ec versus preset current Is of SET devices at16 K and room temperature Labels A B and C and and corre-spond to those in Figures 2 3 and 4 and Figures 4 and 5 respectively

to 190 meV even at 16 K with increasing the preset cur-rent from 1 A to 70 A respectively Figure 6 showsthe charging energy Ec of the devices as a function ofthe preset current Is during the activation The chargingenergy of the devices strongly relates with the preset cur-rent of the activation procedure rather than the initial gapseparation of the nanogaps The results imply that thegrowth of Ni islands with increasing the preset currentcauses the increase of the size of the islands and results inthe decrease of the number of the islands Consequentlythe preset current Is in the activation procedure can con-trol the charging energy and the number of islands dur-ing the formation of the SETs Therefore it is consideredthat initial gap separation and preset current are impor-tant parameters so as to optimize the fabrication schemeof planar-type SETs by activation method

4 CONCLUSIONS

In conclusion we present a novel technique for the fabrica-tion of planar-type Ni-based SETs consisting of nanogapsThis method is based on electromigration induced byF-N field emission current and is called ldquoactivationrdquoWhen the activation was applied to the nanogaps theIdminusVd characteristics of the nanogaps at room temper-ature displayed Coulomb blockade properties Addition-ally Coulomb blockade voltage was modulated by the

gate voltage quasi-periodically resulting in the formationof multiple islands of SETs Furthermore as the presetcurrent Is was increased the charging energy Ec wasdecreased It is suggested that the control of the chargingenergy Ec is possible with the preset current Is of theactivation In contrast to the previously reported methodthe present method allows us to simplify the fabricationprocess further and control the charging energy and thenumber of the islands of the planar-type Ni-based SETsat room temperature The activation is simple and easymethod for the fabrication of planar-type SETs

References and Notes

1 K Liu Ph Avouris J Bucchignano R Martel S Sun and J MichlAppl Phys Lett 80 865 (2002)

2 Y Nakamura C D Chen and J S Tsai Jpn J Appl Phys35 L1465 (1996)

3 S I Khondaker and Z Yao Appl Phys Lett 81 4613 (2002)4 U C Coskun H Mebrahtu P B Huang J Huang D Sebba

A Biasco A Makarovski A Lazarides T H LaBean andG Finkelstein Appl Phys Lett 93 123101 (2008)

5 W Chen H Ahmed and K Nakazoto Appl Phys Lett 66 3383(1995)

6 A Bezryadin C Dekker and G Schmid Appl Phys Lett 71 1273(1997)

7 D L Klein P L McEuen J E B Katari R Roth and A PAlivisatos Appl Phys Lett 68 2574 (1996)

8 H Park A K L Lim A P Alivisatos J Park and P L McEuenAppl Phys Lett 75 301 (1999)

9 K I Bolotin F Kuemmeth A N Pasupathy and D C Ralph ApplPhys Lett 84 3154 (2004)

10 V A Krupenin A B Zorin M N Savvateev D E Presnov andJ Niemeyer J Appl Phys 90 2411 (2001)

11 X Luo A O Orlov and G L Snider Microelectronics J 36 308(2005)

12 S Kayashima K Takahashi M Motoyama and J Shirakashi JpnJ Appl Phys 46 L907 (2007)

13 S Kayashima K Takahashi M Motoyama and J ShirakashiJ Phys Conf Ser 100 052022 (2008)

14 Y Tomoda K Takahashi M Hanada W Kume and J ShirakashiJ Vac Sci Technol B 27 813 (2009)

15 A Anaya A L Korotkov M Bowman J Waddell andD Davidovic J Appl Phys 93 3501 (2003)

16 A Bramanti G Maruccio P Visconti S DrsquoAmico R Cingolaniand R Rinaldi IEEE Trans Electron Devices 53 2958(2006)

17 J Shirakashi K Matsumoto N Miura and M Konagai JpnJ Appl Phys 37 1594 (1998)

18 J Shirakashi K Matsumoto N Miura and M Konagai Appl PhysLett 72 1893 (1998)

19 J Shirakashi K Matsumoto N Miura and M Konagai J ApplPhys 83 5567 (1998)

20 C Dubuc J Beauvais and D Drouin IEEE Trans Nanotechnol7 68 (2008)

Received 4 September 2009 Accepted 30 October 2009



IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE


ldquoactivationrdquo Similar processes have been reported previ-ously in studies of the formation of Au bridges across70 nm slits15 and insulated planar nanoscopic contacts16

These phenomena can be understood on the basis of theelectrical breakdown procedure in planar contacts In thispaper we describe a fabrication scheme of planar-typeNiVacuumNi-based SETs with multiple tunnel junctionsusing activation method Currentndashvoltage (IndashV charac-teristics of the devices formed by the activation clearlyexhibited the suppression of electrical current at low-biasvoltages known as the Coulomb Blockade and Coulombblockade voltage was also modulated by the gate voltagequasi-periodically at room temperature Furthermore weobserved Coulomb staircase at 16 K The control proce-dure of single-electron charging energy and the number ofislands in the devices by the preset current of the activationis studied in detail

2 EXPERIMENTAL DETAILS

Activation experiments were performed on 100 nm thickthermally grown SiO2 surface on Si substrates First25 nm thick Au contact pads were fabricated upon 5 nmTi layer by electron-beam lithography and electron-beamevaporation Then 20 nm thick arrow-shaped Ni nanogapswith separations of 21ndash68 nm were formed by the sameprocedure in the gap between the Au contact pads Thearrow-shaped source-drain electrodes cause the concentra-tion of electric field at the tip of the nanogap electrodesFigure 1(a) shows a scanning electron microscope (SEM)image of the initial nanogap Before performing the acti-vation the nanogap in the device exhibited the separationof 33 nm with high tunnel resistance above 100 TFollowing these fabrication processes we applied the

activation procedure to the nanogaps Initial nanogapswere installed on a stage with manual probes in a vac-uum chamber with a pressure of 10minus3minus10minus4 Pa and wereoptionally selected from among the samples for the acti-vation procedure Through the experiments the activationprocedure was carried out in the vacuum chamber at roomtemperature The electrical properties of the nanogapsduring the activation were controlled and measured by a

Source

Drain

SiO2

Ni

33 nm

50 nm

Ni ClusterDot

lt10 nm

(a) (b)

Source

Drain

Ni

SiO2 50 nm

Fig 1 SEM image of nanogap electrodes (a) before and (b) after acti-vation with the preset current Is = 150 A

semiconductor parameter analyzer The detailed activationprocess for the fabrication of the SETs consists of fourmain steps(1) we set the magnitude of preset current Is(2) the voltage V is applied to the initial nanogap and isramped up while monitoring the current I passing throughthe nanogap(3) we stop the voltage V when the current I reaches apreset current Is and(4) after performing the activation drain current Id -drainvoltage Vd properties of the device are measured with themodulation of gate voltage Vg to confirm SET charac-teristics This procedure was continuously repeated to thedevices with increasing the preset current from 1 nA to150 A

3 RESULTS AND DISCUSSION

The activation procedure was performed for the fabricationof planar-type Ni-based SETs In the fabrication step ofthe SETs as mentioned above the voltage V was appliedto the nanogaps with the preset current Is from 1 nA to150 A Figure 1(a) shows a SEM image of Ni initialnanogap before the activation In the Figure initial gapseparation is estimated to be approximately 33 nm A SEMimage of the nanogap after the activation with the pre-set current Is = 150 A is exhibited in Figure 1(b) Theimage clearly indicates that the gap separation was nar-rowed from 33 nm to less than 10 nm by the accumulationof Ni atoms within the gap suggesting the formation of Niclusterdot structures between source and drain electrodes

102

104

106

108

1010

1012

1014

1016

10ndash10 10ndash9 10ndash8 10ndash7 10ndash6 10ndash5 10ndash4 10ndash3

T = 300 KNi 20 nm

Tunneling Metallic

A

B C

Tun

nel r

esis

tanc

e R

(Ω

)


Insulating

Fig 2 Dependence of tunnel resistances of nanogap electrodes on pre-set current Is The low-voltage tunnel resistance R is measured afterthe activation The separation of the initial gap is about 33 nm LabelsA B and C correspond to the devices activated with the preset currentIs = 400 nA 650 nA and 15 A respectively



IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE





ndash300

ndash200

ndash100

0

100

200

300

05

Dra

in c

urre

nt I d

(fA

)

T = 300 K Is = 400 nANi 20 nm


∆Vg = 01 V step

Sample A(a)


ndash05

0

05

1

ndash35ndash50

ndash25

0

25

50

Dra

in v

olta

ge V

d (V

)

Gate voltage Vg (V)

Dra

in c

urre

nt I d

(fA

)


ndash1 ndash05 0

ndash3

ndash2

ndash1

0

1

2

3


T = 300 KNi 20 nmIs = 650 nA

Sample B(b)

Dra

in c

urre

nt I d

(pA

)

Vg = ndash40~40 V

∆Vg = 5 V step

ndash05

ndash025

0

025

05

ndash2

ndash1

0

1

2

Gate voltage Vg (V)

Dra

in c

urre

nt I d

(pA

)

Dra

in v

olta

ge V

d (V

)


ndash150

ndash100

ndash50

0

50

100

150



Sample C(c)

Dra

in c

urre

nt I d

(pA

)


ndash05

ndash025

0

025

05

ndash30

ndash15

0

15

30

Dra

in v

olta

ge V

d (V

)

Gate voltage Vg (V)

Dra

in c

urre

nt I d

(pA

)





IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE




Initi

al g

ap s

epar

atio

n W

(nm

)


10

20

30

40

50

60

70

80


T = 300 KNi 20 nm

α

β

γ

A

B

C


ndash15

ndash1

ndash05

0

05

1

15

050

50

100

150

200

Dra

in c

urre

nt I d

(pA

)


Con

duct

ance

dId

V (

pS)

T = 16 K

Ni 20 nm

Is = 1 microA

(a)

EC = 400 meV

ndash5

ndash25

0

25

5

0

1

2

3

Dra

in c

urre

nt I d

(pA

)

Con

duct

ance

dId

V (

nS)


T = 16 K

Ni 20 nm

Is = 3 microA

EC = 300 meV

(b)

ndash15

ndash10

ndash5

0

5

10

15

0

50

100

150

200

250

Dra

in c

urre

nt I d

(nA

)

Con

duct

ance

dId

V (

nS)


T = 16 K

Ni 20 nm

Is = 70 microA

EC = 190 meV

(c)

Sample β

Sample γ




Sample α




IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE


102

103

104



T = 300 K W = 21 nm W = 26 nm W = 33 nm

T = 16 K




α β γ

AB

C

Cha

rgin

g en

ergy

Ec

(meV

)



4 CONCLUSIONS



























IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE





ndash300

ndash200

ndash100

0

100

200

300

05

Dra

in c

urre

nt I d

(fA

)

T = 300 K Is = 400 nANi 20 nm


∆Vg = 01 V step

Sample A(a)


ndash05

0

05

1

ndash35ndash50

ndash25

0

25

50

Dra

in v

olta

ge V

d (V

)

Gate voltage Vg (V)

Dra

in c

urre

nt I d

(fA

)


ndash1 ndash05 0

ndash3

ndash2

ndash1

0

1

2

3


T = 300 KNi 20 nmIs = 650 nA

Sample B(b)

Dra

in c

urre

nt I d

(pA

)

Vg = ndash40~40 V

∆Vg = 5 V step

ndash05

ndash025

0

025

05

ndash2

ndash1

0

1

2

Gate voltage Vg (V)

Dra

in c

urre

nt I d

(pA

)

Dra

in v

olta

ge V

d (V

)


ndash150

ndash100

ndash50

0

50

100

150



Sample C(c)

Dra

in c

urre

nt I d

(pA

)


ndash05

ndash025

0

025

05

ndash30

ndash15

0

15

30

Dra

in v

olta

ge V

d (V

)

Gate voltage Vg (V)

Dra

in c

urre

nt I d

(pA

)





IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE




Initi

al g

ap s

epar

atio

n W

(nm

)


10

20

30

40

50

60

70

80


T = 300 KNi 20 nm

α

β

γ

A

B

C


ndash15

ndash1

ndash05

0

05

1

15

050

50

100

150

200

Dra

in c

urre

nt I d

(pA

)


Con

duct

ance

dId

V (

pS)

T = 16 K

Ni 20 nm

Is = 1 microA

(a)

EC = 400 meV

ndash5

ndash25

0

25

5

0

1

2

3

Dra

in c

urre

nt I d

(pA

)

Con

duct

ance

dId

V (

nS)


T = 16 K

Ni 20 nm

Is = 3 microA

EC = 300 meV

(b)

ndash15

ndash10

ndash5

0

5

10

15

0

50

100

150

200

250

Dra

in c

urre

nt I d

(nA

)

Con

duct

ance

dId

V (

nS)


T = 16 K

Ni 20 nm

Is = 70 microA

EC = 190 meV

(c)

Sample β

Sample γ




Sample α




IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE


102

103

104



T = 300 K W = 21 nm W = 26 nm W = 33 nm

T = 16 K




α β γ

AB

C

Cha

rgin

g en

ergy

Ec

(meV

)



4 CONCLUSIONS



























IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE




Initi

al g

ap s

epar

atio

n W

(nm

)


10

20

30

40

50

60

70

80


T = 300 KNi 20 nm

α

β

γ

A

B

C


ndash15

ndash1

ndash05

0

05

1

15

050

50

100

150

200

Dra

in c

urre

nt I d

(pA

)


Con

duct

ance

dId

V (

pS)

T = 16 K

Ni 20 nm

Is = 1 microA

(a)

EC = 400 meV

ndash5

ndash25

0

25

5

0

1

2

3

Dra

in c

urre

nt I d

(pA

)

Con

duct

ance

dId

V (

nS)


T = 16 K

Ni 20 nm

Is = 3 microA

EC = 300 meV

(b)

ndash15

ndash10

ndash5

0

5

10

15

0

50

100

150

200

250

Dra

in c

urre

nt I d

(nA

)

Con

duct

ance

dId

V (

nS)


T = 16 K

Ni 20 nm

Is = 70 microA

EC = 190 meV

(c)

Sample β

Sample γ




Sample α




IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE


102

103

104



T = 300 K W = 21 nm W = 26 nm W = 33 nm

T = 16 K




α β γ

AB

C

Cha

rgin

g en

ergy

Ec

(meV

)



4 CONCLUSIONS



























IP 16593109152Mon 09 May 2011 073955

RESEARCH

ARTIC

LE


102

103

104



T = 300 K W = 21 nm W = 26 nm W = 33 nm

T = 16 K




α β γ

AB

C

Cha

rgin

g en

ergy

Ec

(meV

)



4 CONCLUSIONS


























Fabrication of Single-Electron Transistors Using Field

Documents

Transcript of Fabrication of Single-Electron Transistors Using Field