• No results found

Solid State Devices. Section 20 PN Diode I-V Characteristics

N/A
N/A
Protected

Academic year: 2021

Share "Solid State Devices. Section 20 PN Diode I-V Characteristics"

Copied!
30
0
0

Loading.... (view fulltext now)

Full text

(1)

Klimeck – Solid State Devices

Section 20

PN Diode I-V Characteristics

Gerhard Klimeck

[email protected]

Solid State Devices

(2)

Section 20

PN Diode I-V Characteristics

• 20.1 Band diagram with applied bias

• 20.2 Derivation of the forward bias formula • 20.3 Forward Bias - Non-linear Regime

»Resistive drop »Ambipolar regime • 20.4 Non-ideal effects: »Junction recombination »Tunneling »Impact ionization status V V V V ( ) 2 − β 2 β

=

n p

=

qVA i i F F

n e

n e

n

p

(3)

Klimeck – Solid State Devices (4,6) Junction Recombination ( ) 2 − β 2 β

=

n p

=

qVA i i F F

n e

n e

n

p

(4)

(4,6) Junction Recombination 0 W R n I qA dx t ∂ = − ∂

2 1 1 ( ) ( [ ] [ ] ) ] ) [ ( ( ) τ τ − ∂ = − ∂ + + + i p n p x p x n x n n x n n t p p n

τ

τ

=

E

i

=

E

T

n

1

=

p

1

=

n

i ] 2 ) ( ) ( [ ) 1 ( / 2 i kT qV i n x p x n e n t n A + + − − = ∂ ∂ τ Assume

What is the recombination current?

Shockley-Reed Hall

Follows from assuming midgap traps ( ) 2 − β 2 β

=

n p

=

qVA i i F F

n e

n e

n

p

(5)

Klimeck – Solid State Devices

Electron/Hole Concentrations at Junction

) ( ) (x E qV x Ei = iL − ( ) [ ( ] ) ( ) / / ( ) N i N iL kT F E x F E qV x i kT i n e x n e n − − + = = / 2 [ ]/ [ ( )]/ / ( ) ( ) N A N iL iL A qV kT i kT i F E F E qV qV x kT qV kT i x n e x e n p n e− − + + − + = = position ] 2 ) ( ) ( [ ) 1 ( / 2 i kT qV i n x p x n e n t n A + + − − = ∂ ∂ τ

(6)

Junction Recombination kT E F UFN = NiL q kT V UA A / = ] [ ) 1 ( A FN FN A U U U U U U i e e e n t n + − − + + − − = ∂ ∂ τ ] [ ) ( 2 / 2 / 2 / 2 / 2 / 2 / A FN A FN A A A A U U U U U U U U U U i e e e e e e n t n + − − − + − + − − = ∂ ∂ ⇒ τ ( / 2) [ / 2] i A FN A n sinh U n t τ cosh U U U ∂ ⇒ = − ∂ + −     = −  ni × UA ×

W dx I qA sinh 0 W R n I qA dx t ∂ = − ∂

( ) [ ( ] ) ( ) / / ( ) N i N iL kT F E x F E qV x i kT i n e x n e n − − + = = / 2 [ ]/ [ ( )]/ / ( ) ( ) N A N iL iL A qV kT i kT i F E F E qV qV x kT qV kT i x n e x e n p n e− − + + − + = = ] 2 ) ( ) ( [ ) 1 ( / 2 i kT qV i n x p x n e n t n A + + − − = ∂ ∂ τ

(7)

Klimeck – Solid State Devices

Junction Recombination in Forward Bias

( / 2) 0 ( ) ( ) 2 FN U A W i A R U U n U dx I qA sinh e τ + − ⇒ ≈ −

max ( ) /( / 2 ) 0 ( ) ( ) 2 W i A R x kT q n U dx I qA sinh e τ − ⇒ ≈ − ×

E / 2 max 2 A qV kT i Dep n kT I qA e q τ    ⇒ = −      E  VA 2 4 5 1 3 6,7 ln(I) Emax

Effective spatial width of recombination

Excess Carrier at mid-junction Factor ½ indicates

recombination / traps in junction Traps hidden in τ 0 2 [ / 2] τ     = −  × × + −    

W i A R FN A n U dx I qA sinh cosh U U U

Keep positive branch

Approximate electric field as constant

(8)

Junction Leakage in Practice n p d rj rj Insulating Layer

Junction Design Considerations Electric field stronger at corners, sharp edges.  increased recombination!

(9)

Klimeck – Solid State Devices

Junction Recombination in Reverse Bias

2

i

n

n

t

= −

τ

W=xn+xp (Recombination in depletion region n=p=0) Depletion width 2 1 1 ( ) ( [ ] [ ] ) ] ) [ ( ( ) τ τ − ∂ = − ∂ + + + i p n p x p x n x n n x n n t p p n

τ

τ

=

E

i

=

E

T

n

1

=

p

1

=

n

i ] 2 ) ( ) ( [ ) 1 ( / 2 i kT qV i n x p x n e n t n A + + − − = ∂ ∂ τ Assume Shockley-Reed Hall

Follows from assuming midgap traps ( ) 2 − β 2 β

=

n p

=

qVA i i F F

n e

n e

n

p

(10)

Junction Recombination in Reverse Bias 0

2

R W i

n

I

qA

dx

τ

≈ −

2

i

n

n

t

= −

τ

VA 2 4 5 1 3 6,7 ln(I

)

2

i bi A

W

n

q

A

V

V

τ

= −

W=xn+xp (Recombination in depletion region n=p=0) Integrate… Depletion width

Ideally the flat – now it depends as a square root of voltage => another probe mechanism for traps in junction

(11)

Klimeck – Solid State Devices

Section 20

PN Diode I-V Characteristics

• 20.1 Band diagram with applied bias

• 20.2 Derivation of the forward bias formula • 20.3 Forward Bias - Non-linear Regime

»Resistive drop »Ambipolar regime • 20.4 Non-ideal effects: »Junction recombination »Tunneling »Impact ionization status V V V V

(12)

Section 20

PN Diode I-V Characteristics

• 20.1 Band diagram with applied bias

• 20.2 Derivation of the forward bias formula • 20.3 Forward Bias - Non-linear Regime

»Resistive drop »Ambipolar regime • 20.4 Non-ideal effects: »Junction recombination »Tunneling »Impact ionization status V V V V

(13)

Klimeck – Solid State Devices

Forward Bias Nonlinearity (7): Esaki Diode

I VA Fn F p Fn Fn Fp Fp X

V

A 2 1 3 6,7

ln(I)

Heavy doping 1 empty No states!

Esaki-Diode: Heavily doped diode

Tunneling in diodes. Nobel Prize (Esaki)

(14)

Reverse Bias (5): Zener Tunneling 2 2 4 4 cosh sinh (p.49 ADF) = υ = α   α + α α   I qpT T k d d k Fn Fp 4 5 +VA empty empty Tunneling (triangular barrier)

Remember: Tunneling through a triangular barrier Zener tunneling occurs in every diode. (reverse bias)

(15)

Klimeck – Solid State Devices

Various Regions of I-V Characteristics

1. Diffusion limited 2. Ambipolar transport 3. High injection 4. R-G in depletion 5. Breakdown 6. Trap-assisted R-G 7. Esaki Tunneling

V

A 2 4 5 1 3 6,7

ln(I)

(16)

Section 20

PN Diode I-V Characteristics

• 20.1 Band diagram with applied bias

• 20.2 Derivation of the forward bias formula • 20.3 Forward Bias - Non-linear Regime

»Resistive drop

»Ambipolar regime

• 20.4 Non-ideal effects: »Junction recombination »Tunneling

»Impact ionization status

V

V

V

(17)

Klimeck – Solid State Devices Avalanche Breakdown 1. Diffusion limited 2. Ambipolar transport 3. High injection 4. R-G in depletion 5. Breakdown 6. Trap-assisted R-G 7. Esaki Tunneling

V

A 2 4 5 1 3 6,7

ln(I)

(18)

Nonlinearity due to Impact-Ionization

(19)

Klimeck – Solid State Devices

Nonlinearity due to Impact-Ionization

Exponential current growth

(Impact Ionization or Inverse Auger process)

High Reverse Bias > bandgap (typically 3/2 bandgap)

(20)

Impact-ionization and Flux Conservation ( ) n( ) n( ) p p( ) n n I x + dx = I x + α I x dx + α I x dx 𝑑𝑑𝐼𝐼𝑛𝑛(𝑥𝑥) 𝑑𝑑𝑥𝑥 = 𝛼𝛼𝑝𝑝 𝐼𝐼𝑇𝑇 − 𝐼𝐼𝑛𝑛(𝑥𝑥) + 𝛼𝛼𝑛𝑛𝐼𝐼𝑛𝑛(𝑥𝑥) x W 0

Impact Ionization probabilities

Steady state: Define IT= IN+IP (total current) Constant in steady state

Differential equation 𝑑𝑑𝐼𝐼𝑛𝑛(𝑥𝑥) 𝑑𝑑𝑥𝑥 − 𝛼𝛼𝑛𝑛 − 𝛼𝛼𝑝𝑝 𝐼𝐼𝑛𝑛(𝑥𝑥) = 𝛼𝛼𝑝𝑝𝐼𝐼𝑇𝑇 𝐼𝐼𝑝𝑝 = 𝐼𝐼𝑇𝑇 − 𝐼𝐼𝑛𝑛(𝑥𝑥) ( ) ( ) ( ) ( ) ( ) ( ) ( ) n n n n p p n n n p p I x dx I x I x I x dx dI x I x I x dx + − = α + α ⇒ = α + α

(21)

Klimeck – Solid State Devices

Impact-ionization and Flux Conservation

Differential equation 𝑑𝑑𝐼𝐼𝑛𝑛(𝑥𝑥) 𝑑𝑑𝑥𝑥 − 𝛼𝛼𝑛𝑛 − 𝛼𝛼𝑝𝑝 𝐼𝐼𝑛𝑛(𝑥𝑥) = 𝛼𝛼𝑝𝑝𝐼𝐼𝑇𝑇 𝑦𝑦𝑦 − 𝑎𝑎𝑦𝑦 = 𝑏𝑏 𝑦𝑦 = 1𝑎𝑎 𝑒𝑒𝑎𝑎𝑎𝑎+𝑐𝑐𝑎𝑎 − 𝑎𝑎𝑏𝑏 x W 0 Ip(W)~0 In(0)~0 ( )

(

)

( ) 0 0 ' 0 ' 0 (0) 1 ( ) x n p x n p W dx p T n W dx T p n n e dx I I W I e dx I − α −α − α −α ∫ α + = ∫ + α − α

Solution form of differential equation Reverse diffusion current

(22)

Impact-ionization ( ) ( (0) 1 ( ) n T ) T p T n n p I W I I W I I M I W I + = ⇒ ≈ ≡ � 𝑊𝑊 𝛼𝛼𝑛𝑛𝑒𝑒− ∫0𝑥𝑥 𝛼𝛼𝑛𝑛−𝛼𝛼𝑝𝑝 𝑑𝑑𝑎𝑎𝑑𝑑𝑑𝑥𝑥 = 1 − 1 𝑀𝑀𝑝𝑝 ( )

(

)

( ) 0 0 ' 0 ' 0 (0) 1 ( ) x n p x n p W dx p T n W dx T p n n e dx I I W I e dx I − α −α − α −α ∫ α + = ∫ + α − α

x W 0 Ip(W)~0

In(0)~0 Solution form of differential equation Reverse diffusion current

At x=W, INhas grown exponentially, and IPis now negligible. Multiplication Factor 1 = ∫0 𝑊𝑊𝛼𝛼 𝑝𝑝𝑒𝑒− ∫0 𝑥𝑥 𝛼𝛼 𝑛𝑛−𝛼𝛼𝑝𝑝 𝑑𝑑𝑎𝑎𝑑𝑑𝑑𝑥𝑥 + 1 𝑀𝑀𝑝𝑝 1 + ∫0𝑊𝑊 𝛼𝛼𝑝𝑝 − 𝛼𝛼𝑛𝑛 𝑒𝑒− ∫0𝑥𝑥 𝛼𝛼𝑛𝑛−𝛼𝛼𝑝𝑝 𝑑𝑑𝑎𝑎𝑑𝑑𝑑𝑥𝑥 1 + � 0 𝑊𝑊 𝛼𝛼𝑝𝑝 − 𝛼𝛼𝑛𝑛 𝑒𝑒− ∫0 𝑥𝑥 𝛼𝛼 𝑛𝑛−𝛼𝛼𝑝𝑝 𝑑𝑑𝑎𝑎𝑑𝑑𝑑𝑥𝑥 = � 0 𝑊𝑊 𝛼𝛼𝑝𝑝𝑒𝑒− ∫0 𝑥𝑥 𝛼𝛼 𝑛𝑛−𝛼𝛼𝑝𝑝 𝑑𝑑𝑎𝑎𝑑𝑑𝑑𝑥𝑥 + 1 𝑀𝑀𝑝𝑝 ≈ 1

(23)

Klimeck – Solid State Devices Impact-ionization � 0 𝑊𝑊 𝛼𝛼𝑛𝑛𝑒𝑒− ∫0𝑥𝑥 𝛼𝛼𝑛𝑛−𝛼𝛼𝑝𝑝 𝑑𝑑𝑎𝑎𝑑𝑑𝑑𝑥𝑥 = 1 − 1 𝑀𝑀𝑝𝑝 x W 0 ≈ 1 Assume: 𝛼𝛼𝑛𝑛 = 𝛼𝛼𝑝𝑝 � 0 𝑊𝑊 𝛼𝛼𝑛𝑛 𝑥𝑥 𝑑𝑑𝑥𝑥 = 1

this cannot be quite right… n, p different masses etc

Electric Field

Position

VA<0 VA=0

VA>0

Assume: 𝛼𝛼𝑛𝑛(𝑥𝑥) = 𝛼𝛼𝑛𝑛 Most avalanche happens spatially at maximum electric field 𝛼𝛼𝑛𝑛 𝑊𝑊 = 1

0

B

p A e

α = E from experiment and theory A and B material coeffients

( )

(

)

1 / 2 0 0 2 0 D n D A bi A s s D A qN x q N N V V k k N N= =   +    ε ε E Breakdown-Field

(24)

Impact-ionization: In Practice

Good …. Imagers

single photon can create an avalanche

Bad…. n p Insulating Layer d rj rj Photon Detector

High E-fields at junction corners  Breakdown

(25)

Klimeck – Solid State Devices

Junction Engineering

E E

Reduced field for p-i-n junction, because Vbi (area under the curve) must be the same.

n p Insulating Layer d rj rj i-region Lower E-field! Intrinsic region: E-field has to be constant, because there are no charges!

High practical relevance for MOSFET scaling Needed to reduce fields on the drain side….

(26)

Modern Considerations: Dead Space

For very small (ballistic) junctions,

electrons can cross the junction without inducing impact Ionization.

(Dead space too small)

W

Dead Space

Dead Space:

Space you need before an electron can impact ionize.

(27)

Klimeck – Solid State Devices

Zener Breakdown vs. Impact Ionization

How do you differentiate between Zener tunneling and

impact-ionization? Need bias > bandgap

Can happen for smaller reverse bias ~1V Very noisy – can be measured

(28)

Non-Ideal Effects: Conclusion

1) Junction recombination is often used as a diagnostic tool for process maturity. Defects in junction arises from misplaced donor impurities, not necessary from deep-trap impurities.

2) Impact ionization plays an important role in wide variety of devices (e.g. avalanche photo-diodes).

(29)

Klimeck – Solid State Devices

Section 20

PN Diode I-V Characteristics

• 20.1 Band diagram with applied bias

• 20.2 Derivation of the forward bias formula • 20.3 Forward Bias - Non-linear Regime

»Resistive drop

»Ambipolar regime

• 20.4 Non-ideal effects: »Junction recombination »Tunneling

»Impact ionization status

V

V

V

(30)

Section 20

PN Diode I-V Characteristics

• 20.1 Band diagram with applied bias

• 20.2 Derivation of the forward bias formula • 20.3 Forward Bias - Non-linear Regime

»Resistive drop »Ambipolar regime • 20.4 Non-ideal effects: »Junction recombination »Tunneling »Impact ionization V V V V

References

Related documents

We consider as an important tool for the development of the global educational environment: world trends in the formation of informational educational environment in the

Though the early church practiced almsgiving as a normative act of piety its role as a formative spiritual discipline has been largely ignored in contemporary evangelical works on

In our model, due to the unemployment rate di¤erential across skill groups within a region and across regions for a given skill group, the incentives to acquire human capital and

within the plume. Changes in cloud properties detected by MODIS AQUA for October 2014. Stippling in a) and c) represent areas of 95% 513. confidence level significant

While the parent-group treatment led to significantly greater reductions in chil- dren’s BMI z-scores compared to the standard treatment [ 12 ], children with a higher degree of

induced IL-10 production by F4/80 + CD11b + CD11c int intestinal macrophages, but not CD4 + T cells, in inflamed mucosa, thereby suppressing experimental colitis by an

The specific objectives were to assess the effect of drying schedule on: (i) surface check total length; (ii) honeycomb checks expressed as a percentage of

• The total ground floor area of all detached accessory structures located on a single residential property in the R-1, and R-2 Use Districts shall not exceed 1,000 square feet or