Additive Printing of Flexible Electronics for Sensing

Tse Nga (Tina) Ng Electrical and Computer Engineering Dept., UC San Diego

tnn046@ucsd.edu

1 um

1 mm

Advantages and limitations of printing

resist

Deposit film

Photolithography

resist

Deposit film

mask

Stamp printingDigital Inkjet

• Low-temperature process, low ink consumption

(<10L to cover a 4”wafer)

• Trade-off in resolution and printing speed

inkjet: ~35 micron resolution, web speed ~5miles/hr;

Imprint ~nm resolution, 20 eight-inch wafers/hr

AKT PECVD PolyICVacuum deposition Solution printing in air

3

Applications for printed sensors

• Scalable to large-area, flexible, tunable materials

Human-computer interface (touch, imager, etc.)

• Multi-component arrays that increase selectivity

Low-cost, high-volume for distributed sensing

GE-Avery Dennison

PARC ARPA-E MONITOR

APL 92 (2008) 213303

Someya, U Tokyo

4

1.E+08

1.E+10

1.E+12

600 1000 1400

D*

[Jo

nes

]

wavelength [nm]

Solution processed sensors comparable to conventional Ge

Tunable organic materials with infrared detectivity

comparable to commercial Newport Ge diode

ACS Appl. Mater. Interface 9 (2017) 1654.

Polymer Science, DOI: 10.1039/c7py00241f

Ge photodiode

5

Printed TFTs for local sensor control

M x N lines, interconnect takes

more space than sensors

TFT

With TFTs, M + N lines only

Simple signal conditioning/processing

Alex Blau

Gain/ threshold

detect

sensors

TFT integrated

circuits provides

signal conditioning

before Si chip

6

Key challenge for integrated TFT circuits

Challenge for implementation: Designs that tolerate variations in OTFTs

Monte Carlo Simulation for 100 samples

-for a gain + latch circuit with 7 TFTs

TFT variation

(1s std dev)

Yield

25% 50%

10% 80%

5% 98%

0

5

10

15

20

0 10 20

outp

ut

[V]

input [v]

Sci. Rep (2015) 5, 13457.

Proc. IEEE (2015) 103, 607

• Variation leads to circuit error

• Controlling variation is key to

practical yield

7

Courtesy of Murmann group at Stanford

Similar level of variations: main source of variation is semiconductor,

less impact from channel W/L

0

2

4

6

8

10

12

-10

0

-50 0

50

10

0

Co

un

t

Bin [% deviation from mean current]

p

n

All-printed OTFTs Photolithographically patterned OTFTs

Printed vs photolithographic OTFTs

Sci. Rep (2012) 2, 585.

8

L

WVVVV

pn

pnn

T

p

TDD

inv

,

1

Modify channel surface to adjust VT

• Important to control threshold voltage VT

• Back-channel interface affects VT: electronic dipole, film morphology, etc.

J. Appl. Phys. 113 (2013) 094506

1.0E-12

1.0E-10

1.0E-08

1.0E-06

-15 0 15 30

dra

in c

urr

ent

[A]

gate bias [V]

VT

Gate

Gate insulator

DrainSource

Gate

Gate insulator

DrainSource

(a) after illumination without gate bias

(b) after illumination with negative gate bias

fixed positive

charges in the bulk

sheet charge fixed

at front channel

charges fixedin the bulk

charges fixed in the bulkand at the back channel

Bare substrate

Gate

Gate insulator

DrainSource

Gate

Gate insulator

DrainSource

(a) after illumination without gate bias

(b) after illumination with negative gate bias

fixed positive

charges in the bulk

sheet charge fixed

at front channel

charges fixedin the bulk

charges fixed in the bulkand at the back channel

9

Small molecules:

Suppressing thermal disorder

by side chain location

J. Anthony, Sirringhaus, et al.

Nature Communications 7,

10736 (2016)

Material structures that reduce disorder

McCoullouch, et al.

Nature 515, (2014) 384–388

Polymers:

Reduce tail states by rigid

backbone that reduces torsion

10

Reduced variations in printed OTFTs

1s =8-10%

Mobility =0.6cm2/Vs

1s =12%

Mobility =0.7cm2/Vs

1s =40%

Mobility =0.12cm2/Vs

Uniformity can be improved in both polymer and small molecules

P-type gen 1 P-type gen 2 N-type gen 2

11

From materials to circuit fabrication

1 mm

“S”

“F”

Design, simulate circuitsDevelop ink & devices Build device models

Print and test

IEEE Elec. Dev. Lett. (2013) 34, 271.

12

Other examples of printed circuits

1 mm

Shift register

IEEE Elec. Dev. Lett.,

34 (2013) 271.

Voltage multiplier

Flexible Printed Electronics

(2016) 1, 015002.

1st

stage

5th

Temperature dose tag

Sci Rep (2015), 5, 13457.

13

Desirable to digitize signal near sensor

Amplitude signal prone to attenuation error; frequency signal more reliable

Same freq as before,

will get same readout

Attenuation affects

amplitude measurement

Need to add digitizing circuit near sensor

14

Voltage-controlled oscillator

-6

-3

0

3

6

9

0 50 100

Am

plit

ud

e [

V]

Time [ms]

10Vdd

-6

-3

0

3

6

9

0 50 100

Am

plit

ud

e [

V]

Time [ms]

8Vdd

0

5

10

15

20

25

0 10 20 30

osc

illat

ion

cyc

le t

ime

[m

s]

supply voltage Vdd [V]

15

Using printed components to mimic

skin mechanoreceptor

PARC | 15

Science (2015) 350, 313.

In collaboration with Zhenan Bao group

16

Need to augment spasticity diagnosis

Spatiscity -involuntary activation of muscle, very common in patient with neurological disorders such as stroke, traumatic brain injury, cerebral palsy, etc. affect 764K in US; 17M world wide

UI Chicago

Score Modified Ashworth Scale (MAS)

0 No increase in muscle tone

1 Slight increase in muscle tone, with a catch and

release at the end of the range of motion (ROM)

1+ Slight increase in muscle tone, followed by minimal

resistance throughout the remainder of ROM

2 More marked increase in muscle tone through

most of the ROM, but affected parts easily moved

3 Considerable increase in muscle tone, passive

movement difficult

4 Affected part is rigid in both flexion and extension

17

Issue with reliability in MAS ratings

• 5 patients and 12 tasks: each doctor gave 60 MAS ratings

• Only 27% of the ratings were the same; poor inter-rating reliability, yet dosage is based on this rating

Two doctors’ MAS ratings on the same patients

In collaboration with Dr. Garudadri (Calit2) and Dr. Skalsky at UCSD School of Medicine

18

Prototype glove to quantify spasticity

Glove worn by the doctor during assessment:-measure force (printed pressure sensor by Tekscan) and angular velocity (gyroscope)-Power to move a limb P=F*v

19

Mock patient to calibrate sensor glove

-calibrate sensor glove with a mock patient with changeable resistance (2-20kg)• load cell to measure force • potentiometer to measure angular velocity• the power P=F*v to move the mock limb is recorded

Forc

e [g

]A

ngu

lar

velo

city

[d

eg/s

]

Time (Sampling at 80 Hz)

20

Better resolution than MAS scale

• Quantitative glove measurement allows comparison between rating trials, less dependence on rater perception

• Glove sensor improves the resolution of the spasticity assessment

Pow

er m

easu

red

by

glo

ve [

Nm

/s]

Power setting on mock patient [Nm/s]Po

wer

mea

sure

d b

y gl

ove

[N

m/s

]Power setting on mock patient [Nm/s]

In press, IEEE-NIH 2016 HI-POCT Proceeding

21

Summary

• Apply additive printing to demonstrate organic TFT circuits

1. increase tolerance to device variation issues

2. integrated local digitizing circuits near sensors

• Example application of printed pressure sensor to achieve quantitative

assessment in spasticity diagnosis

1 mm

3.86 x 1.66 mm

22

Acknowledgment

Collaborators:

PARC colleagues

Antonio Facchetti, Northwestern

Zhenan Bao, Stanford

Iain McCollouch, Imperial/KAUST

Funding: UCSD colleagues:

Leanne Chukoskie

Harinath Garudadri

Andrew Skalsky

Michael Yip

Students:

Fei Deng, Padmaja Jonnalagedda,

Zhenghui Wu, Weichuan Yao,

Hyunwoong Kim, Kaiping Wang,

Udit Parekh, Moran Amit