PAI & Halo ion effects on junction activation and leakage€¦ · PAI & Halo ion effects on...

JTG Sept 22, 2010 Halos, lasers & leakage [email protected]

1

PAI & Halo ion effects on junction activation and leakage

presented at IIT10

Michael Current, Current Scientific, San Jose, CA

and

John O. Borland, JOB Technologies, Aiea, Hawaii

John Marino, EAG, East Windsor, NJ

Blake Darby, U Florida, Gainesville, FL

with

Yoji Kawasaki, Renesas, Itami, Japan


2

Process Mix: 4 PAIs, 2 Halos, 0.2 keV BPAI:

Xe: 5, 14 keV

*B36: 4, 20 keV

*In: 5, 14 keV

Ge: 3, 10 keV

*Note: p-type dopant

Halo:

20 keV As

35 keV Sb

Junction:

0.2 keV B, 1e15


3

SRIM Profiles

SRIM08 Profiles

1E+15

1E+16

1E+17

1E+18

1E+19

1E+20

1E+21

1E+22

1E+23

1E+24

1 10 100 1,000

Depth (Angstroms)Conce

ntr

atio

n (

Ato

ms/

cm3)

0.2 kV B20 kV As35 kV Sb20 kV B36B36 recoils14 kV In In recoils14 kV XeXe recoils10 kV GeGe recoils

SRIM08 Profiles

1E+15

1E+16

1E+17

1E+18

1E+19

1E+20

1E+21

1E+22

1E+23

1E+24

0 100 200 300 400 500 600 700

Depth (Angstroms)

Conce

ntr

atio

n (

Ato

ms/

cm3)


B

AsSb

Ge

In, Xe

Ge recoils


4

a-Si LayersEllipsometry: a-Si Thickness

0

2

4

6

8

10

12

14

0 5 10 15 20 25 30 35 40

Atomic Energy (keV)

a-S

i Th

ickn

ess

(n

m)

BGeXeInB36AsSb

Ge

In, Xe

As, Sb

Tox

SRIM08 Profiles

1E+15

1E+16

1E+17

1E+18

1E+19

1E+20

1E+21

1E+22

1E+23

1E+24

0 100 200 300 400 500 600 700

Depth (Angstroms)

Conce

ntr

atio

n (

Ato

ms/

cm3)


B

AsSb

Ge

In, Xe

Ge recoils

Ellipsometry finds only “thin”a-Si layers.

TEM sees no a-Si in the Asand Sb halos.(Halo dose = 3e13 atoms/cm2)


5

Laser Anneals

Laser Anneals:

Ultratech LSA

1220 C

1240 C

1280 C

1320 C

1350 C

Rsheet (Ohm/sq) Jo (uA/cm2)

0.2 keV B, 20 keV B36 “PAI”


6

SIMS: 0.2 keV B and Halos

Xj (As)

Xj (Sb)

B

As

Sb

SRIM08 Profiles

1E+15

1E+16

1E+17

1E+18

1E+19

1E+20

1E+21

1E+22

1E+23

1E+24

0 100 200 300 400 500 600 700

Depth (Angstroms)

Conce

ntr

atio

n (

Ato

ms/

cm3)


B

AsSb

Ge

In, Xe

Ge recoils


7

SIMS: 10 keV Ge, 0.2 keV B & Halos

B Ge

As

Sb


8

SIMS: 14 keV In, 0.2 keV B & Halos

Note: In levels higher than As or Sb for <20 nm. Counter-doping?

As

Sb

InB


9

Activation & Leakage(no Halo)

14 kV Xe

0.2 kV B10 kV Ge, 5 kV Xe

xxxxxxx

“control” = 0.2 keV B with no PAI & no halo

Different PAI ions give very different Rs and leakage


10

RsL Line Scans: Rs & Jo0.2 keV B (and 20 kV B36): no Halo

W9-15

0.1

1.0

10.0

100.0

1,000.0

10,000.0

-100 -50 0 50 100

Position (mm)

Rs(

Ohm

/sq),

Jo(u

A/c

m2)

20 kV B36/0.2 kV B

Laser Scans: 1220 to 1350 C

0.1

1.0

10.0

100.0

1,000.0

10,000.0

100,000.0

-100 -50 0 50 100

Position (mm)

Rs

(Ohm

/sq),

Jo

(uA/c

m2)

0.2 kV B

Rs

Jo

1220 1240 1280 1320 1350

1220 1240 1280 1320 1350

0.2 keV B implants need lasertemps >1240 C to get tominimal leakage.

Addition of B36 “PAI” to the0.2 keV B greatly improvesleakage for lower laser temps.


11

RsL Line Scans: Rs & Jo0.2 keV B (and Ge PAI): no Halo

Deep (10 kV) Ge PAI hashigher leakage than shallow(3 keV) Ge.

Width of “low leakage” stripesare much thinner than goodactivation (Rs), implying intra-scan anneal temperaturevariations.


12

RsL Line Scans: Rs & Jo0.2 keV B (and Ge PAI): no Halo

B activation

B-Ge Doping

100

1,000

10,000

100,000

-100 -50 0 50 100Position (mm)

Rs

(Ohm

/sq)

0.2 kV B3 kV Ge/0.2 kV B10 kV Ge/0.2 kV B

1220 1240 1280 1320 1350

Deeper Ge PAI lowers Bjunction Rs (at the cost ofhigher leakage current).

Implication:

Deeper a-Si layer results inhigher B dopant activation.


13

RsL Line Scans: Rs & Jo0.2 keV B (and Xe PAI): no Halo

?? ?? ??

xxxxxxxxxxxxx xxxxxxXe PAI has much higher leakageand poorer dopant activation thanGe.

Deep (14 kV) Xe worse thanshallow (5 kV) Xe PAI.

Note:

JPV values are very low when Jois >2e-2 A/cm2 (>2e4 uA/cm2) soRs values are questionable.


14

RsL Line Scans: Rs & Jo0.2 keV B (and In PAI): no Halo

W9-21

0.1

1.0

10.0

100.0

1,000.0

10,000.0

100,000.0

-100 -50 0 50 100

Position (mm)

Rs

(Ohm

/sq),

Jo

(uA/c

m2)

14 kV In/0.2 kV B

W9-24

0.1

1.0

10.0

100.0

1,000.0

10,000.0

100,000.0

-100 -50 0 50 100

Position (mm)

Rs

(Ohm

/sq),

Jo(u

A/c

m2)

5 kV In/0.2 kV B

1220 1240 1280 1320 1350

1220 1240 1280 1320 1350

Leakage for 115In “PAI” muchlower than 131Xe or 73Ge.

Possible effect of In PAIincreasing p-doped Xj to deeperthan main residual damage ??

Active In decreases Rsheet athigher temps (see next slide).


15

RsL Line Scans: Rs & Jo0.2 keV B (and In PAI): no Halo

B-In Doping

1,000

10,000

100,000

-100 -50 0 50 100Position (mm)

Rs

(Ohm

/sq)

0.2 kV B5 kV In/0.2 kV B14 kV In/0.2 kV B

1220 1240 1280 1320 1350

At >1280 C, In “PAI” resultsin lower Rsheet.

An activation increase or anXj shift?

For lower (<1280 C) temps,5 keV In raises the Rsheet

above the 0.2 keV B alone.

Implication: Higher tempsneeded to fully activate In.


16

RsL Line Scans: Rs & Jo0.2 keV B (no PAI): As or Sb Halo

Addition of of halo profiles stronglyincreases carrier recombination andtunneling leakage currents.

As halo activation and leakagebetter than Sb, but only at >1280 C.

“Modest” leakage (≈1e-3 A/cm2)only seen in the middle of the 1320and 1350 C scans for As halo.

W61-25

0.1

1.0

10.0

100.0

1,000.0

10,000.0

100,000.0

-100 -50 0 50 100

Position (mm)

Rs

(Ohm

/sq),

Jo (

uA/c

m2)

0.2 kV B/20 kV As

W6-01

0.1

1

10

100

1000

10000

100000

-100 -50 0 50 100

Position (mm)Rs

(Ohm

/sq),

Jo (

uA/c

m2)

0.2 kV B/35 kV Sb

1220 1240 1280 1320 1350

1220 1240 1280 1320 1350

xxxxxxxxxxxxxxxx

xxxxxxxxxx xx xx


17

RsL Line Scans: Rs & Jo0.2 keV B (14 keV In PAI): As or Sb Halo

W9-22

0.1

1.0

10.0

100.0

1,000.0

10,000.0

100,000.0

-100 -50 0 50 100

Position (mm)

Rs

(Ohm

/sq),

Jo (

uA/c

m2)

14 kV In/0.2 kV B/20 kV As

W9-23

0.1

1.0

10.0

100.0

1,000.0

10,000.0

100,000.0

-100 -50 0 50 100

Position (mm)

Rs

(Ohm

/sq),

Jo (

uA/c

m2)

14 kV In/0.2 kV B/35 kV Sb

1220 1240 1280 1320 1350

1220 1240 1280 1320 1350

Deep 14 keV In “PAI” results in lowerleakage currents for for the As and Sbhalos.

Implication:

Active In counter-dopes the upperportions of the halo profiles, increasingdepletion layer thickness and loweringrecombination and tunneling currents.

Also possible deeper junction Xj.


18

Sheet Resistance

Halo and PAI

10

100

1000

10000

-100 -50 0 50 100

Position (mm)

Rs

(Ohm

/sq)

0.2 kV B/ 20 kV As20 kV B36/ 0.2 kV B/ 20 kV As4 kV B36/ 0.2 kV B/ 20 kV As4 kV B36/ 0.2 kV B/ 35 kV Sb14 kV In/ 0.2 kV B/ 20 kV As14 kV In/ 0.2 kV B/ 35 kV Sb10 kV Ge/ 0.2 kV B/ 20 kV As10 kV Ge/ 0.2 kV B/ 35 kV Sb3 kV Ge/ 0.2 kV B/ 20 kV As

0.2 kV B, various PAI, no halo

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

-100 -50 0 50 100Position (mm)

Rs

(Ohm

/sq)

0.2kV B20kV B36/ 0.2kV B4kV B36/ 0.2kV B14kV Xe/ 0.2kV B5kV Xe/ 0.2kV B14kV In/ 0.2kV B5kV In/ 0.2kV B10kV Ge/ 0.2kV B3kV Ge/ 0.2kV B

4PP: As Halo

4PP: no Halo RsL: no Halo

RsL: As & Sb Halo

No Halo:RsL generally lowervalues than 4PP.

As or Sb Halo:4PP values all toolow (due tosubstrate leakage).

For deep In PAI,4PP values valid(modest leakage).


19

PAI & Halo ion effects on junction activation and leakage:Summary

1. The specific ions used for PAI and halo implants greatlyinfluence Rs and leakage values. It is not just “damage”.

2. Halo profiles greatly increase leakage levels and increaseeffect of residual damage from PAI implants. (Not news).

3. “PAI” with p-type ions (B36, In) reduce Boron junction leakagecurrents, especially for halo profiles. Counter-doping?

4. Significant amount of intra-scan temperature variations.

14 kV Xe

0.2 kV B10 kV Ge, 5 kV Xe

xxxxxxxno halo

0.2 keV B, no PAI, no halo

Date post:	07-Oct-2020
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

PAI & Halo ion effects on junction activation and leakage€¦ · PAI & Halo ion effects on...

Documents