A nanoparticle-organic memory field-effect transistor behaving as a programmable

spiking synapse

F. Alibart, S. Pleutin, D. Guerin, K. Lmimouni, D. Vuillaume Molecular Nanostructures & Devices group,

Institute for Electronics Microelectronics and NanotechnologyCNRS and University of Lille (France)

C. Novembre, C. GamratCEA, LIST/LCE (Advanced Computer Technologies and Architectures)

Saclay (France)

1

NABAB project

Objectives• feasibility of nano computing block (NAB) based on molecular

electronics nanoscale devices, which can acquire a specific computing function by means of a post-fabrication adaptation process (learning, reconfiguration). Neuronal computer.

2

• Function : multiplier• Device: transistor with

gain

• Function : synaptic weight• Device: data storage,

memory+ = NOMFET

• scalable down to few tens of nanometer• low cost• suitable for flexible substrate• bottom-up approaches and organic

But not a non-volatile memory.Need a "leaky" memory

synaptic plasticity

Nanoparticle Organic Memory FET

3

L= 200 nm to 20 µm

Si P++SiO2 200nm

Au Au

SH

Si(OMe)3

SHSHSH SH

S

S

SS S

S

S

SS S

2) Immersion in Au dodecanthiol nanoparticles solution (in toluene)

1)

3) Pentacene evaporation35 nm @ 0.1Å/s

5 nm3.7x1011 NP/cm2

10 nm1.8x1011 NP/cm2

20 nm9x1010 NP/cm2

F. Alibart et al., Adv. Func. Mater., in press

Memory behavior

4

C. Novembre et al., Appl. Phys. Let., 92, 103314 (2008)

Positive charge of the NPs = Negative voltage shift

τ = 95 and 430 s

Q = ΔVTxCox

Biological spiking synapse : dynamic behavior

5

• Facilitating or depressing synapse

For a given frequency f, two post-synaptic responses are possible depending on the type of synapse.

• Short-term plasticity The post-synaptic signal

depends on the frequency of the input signal

Time (ms)

200

200 400

1200400 600 800 1000

600 800 1000 12000

10

50

0

1

0,5

0

mV

mV

depression : the weight decreases as function of time

facilitation : the weight increases as function of time

Boegerhausen (2003 )Tsodyks (1998 )

Time (s)In

put

Out

put W

increases

W decreases

Time (s)

Programmable NOMFET synapse

6

PROGRAMMATION•Write pulse = facilitating behaviour (due to discharge of NPs)•Erase pulse = depressing behaviour (due to charge of NPs)

F. Alibart et al., Adv. Func. Mater., in press

short term plasticity

7

Model calculationdevice response

d

b

c

a

P

P

NP = 5 nm / L = 12 µm

NP = 5 nm / L = 2 µm NP = 5 nm / L = 200 nmm

F. Alibart et al., Adv. Func. Mater., in press

short term plasticity

7

Model calculationdevice response

d

b

c

a

P

P

NP = 5 nm / L = 12 µm

NP = 5 nm / L = 2 µm NP = 5 nm / L = 200 nmm

F. Alibart et al., Adv. Func. Mater., in press

8

NPs size decreases

20 nm

5 nm

20 µm 200 nm

Interelectrode gaps decreases

biological synapse - NOMFET analogy

9

Nn neurotransmiter (NT) activated at spike n

Recovery of NT with a time τd

biological synapse - NOMFET analogy

9

Nn neurotransmiter (NT) activated at spike n

Recovery of NT with a time τd

Gold nanoparticles

charge of Nn holes in the NP at spike n

Pentacene thin film

Discharge of the holes with a time τd

Conclusions• the NOMFET behaves as a biological spiking synapse

– it is programmable– STP is working

• it can be srinked down to 200 nm channel length and 5 nm NPs. (work in progress for L< 200 nm)– possibility of high-integration

• we also developed a theoretical model suitable for device and circuit simulation.– see poster : O. Bichler et al, session B

• Integration of several NOMFET in simple neuronal circuits (perceptron, Hopfield network,...) is in progress.

10