Single Event Transient Response of InGaAs MOSFET

Kai Ni1, En Xia Zhang1, Nicholas C. Hooten1, William G. Bennett1, Michael W. McCurdy1, Ronald D. Schrimpf1, Robert A. Reed1,

Daniel M. Fleetwood1, Michael L. Alles1, Tae-Woo Kim2, Jianqiang Lin3,Jesús A. del Alamo3

1Vanderbilt University, Nashville, TN 37235, USA 2SEMATECH

3Massachusetts Institute of Technology, Cambridge, MA 02139 USA

6/23/14 RER group meeting 1

Motivation

• III-V materials are promising channel candidates at beyond 14nm technology node

• Previous transient studies focus on III-V MESFET/HEMT• Gate transients• Charge enhancement due to source drain pathway• Gate bias dependence shows a peak at threshold voltage

• It’s important to study the transient response of III-V MOSFET

6/23/14 RER group meeting 2

Device II

• Cross section and vertical band diagram

6/23/14 RER group meeting 3

0.00 0.02 0.04 0.06 0.08-4

-3

-2

-1

0

1

2

ener

gy (e

V)

position (µm)

Conduction Band Fermi Level Valance Band

Heavy Ion Results• No gate transient due to large barrier• Source and drain transient have the same magnitude• Two processes with different time constant

6/23/14 RER group meeting 4

• Fast collection: τ≈ 100 ps, directcollection

• Slow collection: τ≈ 3 ns, source-to-drain pathway

0 10 20 30 40 50

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

curr

ent (

mA

)

time (ns)

drain source gate

Gate Bias Dependence• Peak drain current reaches a maximum around threshold voltage

• Peak drain current decreases considerably in inversion and slightly indepletion and accumulation

6/23/14 RER group meeting 5

-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.60.24

0.26

0.28

0.30

0.32

0.34

0.36pe

ak d

rain

cur

rent

(mA

)

VGS-VTH (V)

Laser Irradiation

• Laser wavelength 1.26 μm• Photon energy 0.98 eV, larger than the channel material

bandgap, smaller than the other matherials

6/23/14 RER group meeting 6

XX’

40 µm

10 µ

m

4 µm

Laser Results• Line scan

• The drain side strike has a larger peak drain current than the source side strike

• The drain side has a higher electric field than the source side (VDS=0.5 V), so the electron velocity is higher in drain side, the peak drain current is larger

6/23/14 RER group meeting 7

-2 -1 0 1 2

2.0

2.1

2.2

2.3

2.4

2.5

Drain

peak

dra

in c

urre

nt (m

A)

x position (µm)

Source

Gate Bias Dependence• The laser data is consistent with heavy ion data

• The peak drain current is maximum around threshold and decreases considerably in inversion while slightly in depletion and accumulation

6/23/14 RER group meeting 8

-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.61.7

1.8

1.9

2.0

2.1

2.2

2.3

2.4

2.5pe

ak d

rain

cur

rent

(mA

)

VGS-VTH (V)

2D TCAD Simulation• Model

• The LET is 0.1 pC/μm• The ion strike is gaussian both in space and time• The strike center is at 0.2 μm and 1.0 ns

6/23/14 RER group meeting 9

0 0.2 0.4-0.2-0.4

0.2

0

InAlAs

InGaAsInGaAsInGaAs

HfO2

X

X’

Y’Y source drain

gate

X (µm)

Y (µ

m)

2D TCAD Simulation• Hole density (colored map) and electrical potential (contour)

• At the center of the strike, the electric potential is strongly distorted due to a large number of electrons and holes are generated around the strike location

6/23/14 RER group meeting 10

Pre Strike

Hole Density

0.1

0.2

0.3

0

-0.1

-0.2

0 0.1 0.2 0.3-0.1-0.2

0.1

0.2

0.3

0

-0.1

-0.2

0 0.1 0.2 0.3-0.1-0.2

1.0 ns

2D TCAD Simulation• Hole density (colored map) and electrical potential (contour)

• At 1.2 ns, the electric potential in the buffer recovers. Electrons and holes move to the channel layer. Only the channel layer is strongly perturbed

6/23/14 RER group meeting 11

0.1

0.2

0.3

0

-0.1

-0.2

0 0.1 0.2 0.3-0.1-0.2

1.0 ns

0.1

0.2

0.3

0

-0.1

-0.2

0 0.1 0.2 0.3-0.1-0.2

1.2 ns

X (µm)

Y (µ

m)

2D TCAD Simulation• Conduction band along the horizontal cut and vertical cut

• Electric field exists before strike to stop holes entering channel, but it is almost zero after strike. Electrons and holes flood into the channel

• Electric potential is distorted around the strike location at the center of strike (1.0 ns) but recovers quickly

• At pre-strike, the source-channel barrier is 0.52 eV, but reduces to 0.03 eV at 1.2 ns. The device is ON

• The source-channel barrier is slowly recovered, which lasts for a few nanoseconds 6/23/14 RER group meeting 12

-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

cond

uctio

n ba

nd (e

V)

x position (µm)

1.0 ps 1.0 ns 1.2 ns 1.6 ns 2.1 ns 6.0 ns 10.0 ns

-0.01 0.00 0.01 0.02 0.03 0.04

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

1.0 ps 1.0 ns 1.2 ns 1.6 ns 2.1 ns 6.0 ns 10.0 ns

cond

uctio

n ba

nd (e

V)y position (µm)

2D TCAD Simulation• Conduction band and excess electron density at different gate biases

• The voltage drop along the channel region decreases with gate bias, leading to smaller horizontal electric field, hence smaller velocity

• The absolute excess electron density, the difference between post-strike and pre-strike electron density, reaches a maximum around threshold, decreases considerably in inversion and slightly in depletion and accumulation

6/23/14 RER group meeting 13

-0.4 -0.2 0.0 0.2 0.4

-0.80

-0.75

-0.70

-0.65

-0.60

-0.55

-0.50

-0.45

-0.40

cond

uctio

n ba

nd e

nerg

y (e

V)

x position (µm)

VGS=-0.4 V VGS=-0.2 V VGS= 0.2 V VGS= 0.4 V

0.000 0.002 0.004 0.006 0.0080

1x1018

2x1018

3x1018

4x1018

5x1018

6x1018

7x1018

8x1018

9x1018

exce

ss e

lect

ron

dens

ity (c

m-3)

y position (µm)

VGS=-0.4 V VGS=-0.2 V VGS= 0.2 V VGS= 0.4 V

Comparison• Comparison of the gate bias dependence between heavy ion, laser and

2D TCAD simulation

• The heavy ion data, the laser data and 2D TCAD simulation data agree very well

6/23/14 RER group meeting 14

-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.60.60

0.65

0.70

0.75

0.80

0.85

0.90

0.95

1.00

1.05

Laser Results Heavy Ion Results 2D TCAD Simulation

norm

aliz

ed p

eak

curr

ent

VGS-VTH (V)

Conclusion No gate transients due to large barrier for both electron and hole between

gate dielectric and semiconductor The slow holes piling up under the gate and the source access region

modulates the source channel barrier, turning ON the device and enhancing the collected charge

The peak drain current is maximum for gate biases around threshold anddecreases considerably in inversion and slightly in depletion andaccumulation

Depending on the application and the opportunities for remediation,these transient responses may impose limitations on the use of sometypes of alternative-channel materials in space applicaiton

6/23/14 RER group meeting 15