for biosensing and many other applications Frans Widdershoven Smart - - PowerPoint PPT Presentation

for biosensing and many other applications
SMART_READER_LITE
LIVE PREVIEW

for biosensing and many other applications Frans Widdershoven Smart - - PowerPoint PPT Presentation

CMOS Pixelated Capacitive Sensor platform for biosensing and many other applications Frans Widdershoven Smart sensors NEREID workshop 21st October 2016 COVER PAGE SUBTITLE PLACEHOLDER COMPANY CONFIDENTIAL CMOS integration victory: radio


slide-1
SLIDE 1

COVER PAGE SUBTITLE PLACEHOLDER

COMPANY CONFIDENTIAL

CMOS Pixelated Capacitive Sensor platform for biosensing and many other applications

Frans Widdershoven Smart sensors NEREID workshop 21st October 2016

slide-2
SLIDE 2

CMOS integration victory: radio

2.

Vacuum tubes Transistors Bipolar analog radio IC CMOS digital radio IC (software-defined radio)

COMPANY PUBLIC October 21, 2016

slide-3
SLIDE 3

CMOS integration victory: camera

3.

Vacuum tube (Vidicon) Film Charge-coupled device (CCD) CMOS image sensor (CIS)

COMPANY PUBLIC October 21, 2016

slide-4
SLIDE 4

CMOS integration victory: storage

4.

Ticker tape Magnetic tape Hard disk CMOS-based solid- state drive (SSD)

COMPANY PUBLIC October 21, 2016

slide-5
SLIDE 5

Widdershoven’s 3 laws of IoT

COMPANY PUBLIC

5.

October 21, 2016

1) Only non-trivial data need be transmitted 2) Autonomous devices need sensors to generate non-trivial data 3) What can be sensed by CMOS will be sensed by CMOS Examples:

Voltage, current, power, temperature, RF spectrum, ambient light, magnetic field, images, radar, information, uniqueness (Physically Unclonable Functions), … … and this trend continues!

slide-6
SLIDE 6

Extend the CMOS integration victory to biosensing?

COMPANY PUBLIC

6.

October 21, 2016

Maybe, but only if we stay close to “standard” CMOS

  • It’s the result of a multi-B$ world-wide aligned development effort, so don’t mess

it up!

  • Exploit its strengths:
  • Small feature sizes
  • High speed & low power
  • Embedded signal conditioning, A/D conversion, programmability,…
  • Low-cost and high-yield volume production
slide-7
SLIDE 7

Biological length scales

CMOS technology nodes in production (October 2016)

7.

October 21, 2016 COMPANY PUBLIC

slide-8
SLIDE 8

IgG antibody compared to 14-nm finFET

TEM cross-section through 2 fins (Intel)

8.

COMPANY PUBLIC October 21, 2016

slide-9
SLIDE 9

October 21, 2016 COMPANY PUBLIC

9.

“Capacitive” sense electrode

Electrolyte Double layer SAM Electrode

Captured target Probe Linker 𝑎𝐸𝑀 𝑎𝐹 𝐷𝑇 𝜇𝐸

𝜇𝐸: Debye length (~0.8 nm at 150 mM salt concentration) Cross-section: Equivalent circuit:

slide-10
SLIDE 10

COMPANY PUBLIC

10.

Disturbs

October 21, 2016

Electrolyte Double layer SAM Electrode

Configurations of probes and captured targets Non-specific binding and molecular charge

+ −

SAM defects and relaxation

− −

Intrinsic disturbs Extrinsic disturbs

slide-11
SLIDE 11

Sensitivity scaling

COMPANY PUBLIC

11.

𝑊

1

𝑊

2

𝑊 𝑎1 𝑎2 𝐽 𝑎1 = ℎ1 𝐵1𝜏1 ℎ1, 𝐵1, 𝜏1: height, area and complex conductivity of element 1 𝑎 = 𝑎𝑇 + 𝑎𝐶 = 𝑊 𝐽 = 𝑊 𝐵1𝐾1 = 𝑊 𝐵1𝜏1𝐹1 𝑍 = 1 𝑎 ∆𝑍 ≈ ∆𝜏1 𝜖𝑍 𝜖𝜏1 = Ω1∆𝜏1 𝐹1 𝑊

2

Ω1 = ℎ1𝐵1 (volume of element 1)  Sensitivity is proportional to the square of the local electric field strength!

October 21, 2016

slide-12
SLIDE 12

COMPANY PUBLIC

12.

Model system: semi-spherical metal nanoelectrode

October 21, 2016

Relative sensitivity =

𝑠0𝐹𝑠 𝑊 2

𝑠

Electrolyte Electrode

𝐹𝑠

𝑠0 = 85 nm

A B C

𝜇𝐸

Distance above electrode (nm)

slide-13
SLIDE 13

Frequency ranges

COMPANY PUBLIC

13.

Frequency range Bulk sensitivity Surface sensitivity A (< 3.3 MHz) Low (“blocking” double layer) High (saturated at low-frequency level) B (3.3 – 360 MHz) Nominal (“transparent” double layer) High (still exceeding bulk sensitivity) C (> 360 MHz) Nominal (“vanished” double layer) Nominal (same as bulk sensitivity)

𝑔

2 = 𝜏𝐹,𝐸𝐷

2𝜌𝜗𝐹 𝑔

1 ≈

𝑔

2

1 + 𝑠0 𝜇𝐸 Semi-spherical nano electrode (𝑠

0 = 85 nm)

and 150 mM salt concentration: 𝑔

1 ≈ 3.3 MHz, 𝑔 2 = 360 MHz

𝑔

1

𝑔

2

𝑠0 = 85 nm 150 mM

October 21, 2016

slide-14
SLIDE 14

COMPANY PUBLIC

14.

Validation by numerical simulations

Inside double layer Touching SAM surface Above double layer

SAM: 2.5 nm thick; PNA/DNA: 13.2 nm long (40-bp)

Federico Pittino, Federico Passerini, Luca Selmi, Frans Widdershoven, Microelectronics Journal 45 (12), December 2014

October 21, 2016

slide-15
SLIDE 15

Probe molecules

  • Typically much larger than 𝜇𝐸 (e.g. largely extending above double layer)
  • Should not stick directly to SAM surface (to avoid denaturing and to keep capturing sites accessible for

target molecules)

Other molecules

  • May stick directly to SAM surface (non-specific binding)

Issue

  • At frequencies below 𝑔

2 the sensitivity for non-target molecules and/or SAM surface damage is much

higher than for target molecules

Solution

  • Use modulation frequencies ≥ 𝑔

2 (or at least as high as possible)*

* You won’t find this experimentally by searching for the frequency that gives the highest response

Issues

COMPANY PUBLIC

15.

October 21, 2016

slide-16
SLIDE 16

NXP’s CMOS Pixelated Capacitive sensor chip

ΦT ΦD t2 t1 t ΦT ΦD VD C VT t = t1 VL VN ΦT ΦD VD C VT t = t2 VL VN

𝑅 = 𝑀0 𝑊𝑈 − 𝑊

𝐸

𝐷 + 𝐷𝑄 ∆𝑅 = 𝑀0 𝑊𝑈 − 𝑊

𝐸 ∆𝐷

𝑀0: number of charge/discharge cycles 𝐷𝑄: parasitic capacitance (~0.4 fF) Typically operated at 1 – 50 MHz (runs up to 320 MHz)

16.

COMPANY PUBLIC October 21, 2016

slide-17
SLIDE 17

Cumulative reset noise 2 switch transistors: 𝜏𝑅2 = 2𝑂𝑙𝐶𝑈 𝐷 + 𝐷𝑄

… and no 1/f noise (at least in principle)

Intrinsic signal/noise ratio (SNR): 𝑇𝑂𝑆0 = Δ𝑅 2 𝜏𝑅2 = 𝑂 𝑊𝑈 − 𝑊

𝐸 2 ∆𝐷 2

2𝑙𝐶𝑈 𝐷 + 𝐷𝑄

Intrinsic SNR of single sensor cell

COMPANY PUBLIC

17.

October 21, 2016

Reason for using nanoelectrodes Reason for using advanced CMOS

slide-18
SLIDE 18

Cross-section (90-nm CMOS)

p-well poly-gate source/drain VT ΦT nano-electrode moisture barrier metal-2 metal-3 metal-1 via-1 via-2 contact via-4 metal-4 via-3 VD ΦD ΦD ΦT ΦT ΦD ΦD ΦT ΦT ΦT VT VD VT

18.

Contrast layers Nanoelectrode Early example with e-less Au nanoelectrodes (not used anymore) M1 M2 M3 M4 Cu E-less Au TaN/Ta

COMPANY PUBLIC October 21, 2016

slide-19
SLIDE 19

NXP’s CMOS Pixelated Capacitive sensor chip (2)

Chip features:

1) 256×256 (= 65,536) nanoelectrodes 2) 4 temperature sensors 3) 8 A/D converters 4) 256 digital data accumulators

Flip

  • ver

Fluidic seal patch

Fluid ports

Modified CSP test socket

(Aries Electronics part number A1924-314-23)

3.2 mm × 2.1 mm in 90-nm CMOS (TSMC) Current process: Au-rich AuCu nanoelectrodes, made “the CMOS way”

Spring pins

Thermal interface via backside

  • f chip (not shown)

19.

COMPANY PUBLIC October 21, 2016

slide-20
SLIDE 20

Reconfigurable counter electrode

Active row Counter electrode

20.

COMPANY PUBLIC October 21, 2016

slide-21
SLIDE 21

Reconfigurable counter electrode

Active row Counter electrode Counter electrode

21.

COMPANY PUBLIC October 21, 2016

slide-22
SLIDE 22

Reconfigurable counter electrode

Counter electrode Active row Counter electrode

22.

COMPANY PUBLIC October 21, 2016

slide-23
SLIDE 23

Reconfigurable counter electrode

Counter electrode Active row Counter electrode

23.

COMPANY PUBLIC October 21, 2016

slide-24
SLIDE 24

Lower cut-off frequency (region A  B)

24.

COMPANY PUBLIC October 21, 2016

slide-25
SLIDE 25

Quantitative agreement with simulations

25.

COMPANY PUBLIC October 21, 2016

Measured Simulated

slide-26
SLIDE 26

Detection of particle conduction type

26.

Conducting Insulating

COMPANY PUBLIC October 21, 2016

slide-27
SLIDE 27

Detection of nanoparticle binding

27.

  • Nanoparticles captured by BSA layer on chip surface
  • Independent verification with AFM

COMPANY PUBLIC October 21, 2016

Unpublished pictures removed

slide-28
SLIDE 28

Detection of … virus (one of the smallest viruses)

28.

COMPANY PUBLIC October 21, 2016

Unpublished pictures removed

slide-29
SLIDE 29

Collecting statistics of captured nanoparticles

29.

COMPANY PUBLIC October 21, 2016

Unpublished pictures removed

slide-30
SLIDE 30

Counting of particles and imaging of living cells

30.

COMPANY PUBLIC October 21, 2016

1-µm dielectric particles in water (pH = 3) MCF7 breast tumor cells in growth medium

slide-31
SLIDE 31

Imaging of droplets in water-based emulsions

31.

COMPANY PUBLIC October 21, 2016

Unpublished pictures removed

slide-32
SLIDE 32

Adding smartness: automatic particle tracking

32.

COMPANY PUBLIC October 21, 2016

Unpublished pictures removed

slide-33
SLIDE 33

The future: surfing Moore’s Law?

COMPANY PUBLIC

33.

October 21, 2016

40-nm CMOS design exercise

  • 3 times cell area shrink
  • Resolution comparable to

that of optical microscopes

And what about 14-nm CMOS?

slide-34
SLIDE 34

Udine University (I)

  • Luca Selmi, Federico Pittino, Federico Passerini, Pierpaolo Palestri, Andrea Bandiziol,

Paolo Scarbolo, Andrea Cossettini University of Twente (NL)

  • Serge Lemay, Cecilia Laborde, Christophe Renault, Vincent de Boer, Regine van der Hee,

Jeroen Cornelissen Wageningen University & Research (NL)

  • Maarten Jongsma, Harrie Verhoeven

Thanks to my great collaborators!

COMPANY PUBLIC

34.

October 21, 2016

slide-35
SLIDE 35