Alternative Use of Silicon Edoardo Charbon TU Delft Alex Dommann - - PowerPoint PPT Presentation

Alternative Use of Silicon

Edoardo Charbon TU Delft Alex Dommann EMPA Pantelis Georgiou Imperial College London Bruno Murari ST Microelectronics Roland Thewes TU Berlin Chair: Giovanni De Micheli, EPFL

Edoardo Charbon TU Delft

!"#$%&'()$*+,$*-.*/0"01-&* Do we need to understand Quantum Mechanics, fully?

23*45'%6-&*

/-7$*-80&0-&,*9'#*#50,*,:78-,0;7<*

• !"#$%&'())(&*+%)(,%-&.&/,0123,/&/4%0)1&)2"#3&,42&5"/($/&

%6&/$(23$2&"31&237(322#(37&8423&,429&"#2&6#2/4 *:23-;&

• <)52#,%&="37(%>"33(?@(3$23,2))(&*ABC-&.&%32&:0/,&

)2"#3&,42&6031":23,")/&D#/,&"31&,423&2>2#9,4(37& 2)/2;&

• !"#$%&B"/")2&E%//(&*=93%+9/-&.&A3(>2#/(F2/&/4%0)1&

,2"$4&4%8&,%&)2"#3;&

G&HIJK&L1%"#1%&B4"#5%3& K&

=>?@*A-6$"*B%0C$*0&*45$70,#%:*

=,26"3&M;&N2))& 6%#&4(/&$%3,#(50F%3/&,%&=0+2#?E2/%)0F%3&!($#%/$%+9&

!"12&+%//(5)2O&(3&7#2",&+"#,O&59&+4%,%3& $%03F37&$":2#"/&&

D;'&#;7*E$15'&01,*

G&HIJK&L1%"#1%&B4"#5%3& P& =%0#$2Q&R2()&S;&'03,42#&

F*!1#;'"":F*

• T0"3,0:&!2$4"3($/&(/&",&,42&$%#2&%6&2)2$,#%3($/O&50,&

U0"3,0:&2V2$,/&$%0)1&52&:%#2&%#&)2//&(73%#21&03F)& #2$23,)9&&

• LW":+)2/&%6&,#231Q&&

– #;&&$"0&G*H2IO&"&:"X%#&+)"92#&(3&/05?JI3:&B!Y=& – ,0&G"$J85-#-&*')'"'&15$*K0-K$,*9/B!L,<O&"&328&+4%,%3($& /23/%#&6%#&+4%,%3&$%03F37&(3&B!Y=&

G&HIJK&L1%"#1%&B4"#5%3& Z&

/0&G"$J85-#-&*')'"'&15$*K0-K$,*9/B!L,<* .-%*85-#-&*1-;&(&G*

G&HIJK&L1%"#1%&B4"#5%3& [&

p+ n- n+ V +

• depletion

region

Reverse bias

!"#$%&'#

G&HIJK&L1%"#1%&B4"#5%3& \&

+?82))&70"#1&#(37& +]&"3%12& !0)F+)($"F%3& #27(%3& =^_& `%)9&7",2& a#"8(37Q&a;&=,%++"&

a22+&#2,#%7#"12&3?82))&&

4EM/*85-#-&*1-;&(&G*

HIIK&*L`bc-& ^%1"9& *=^!-&

N5:*6-#5$%*O0#5*85-#-&* 1-;&(&GP*

G&HIJK&L1%"#1%&B4"#5%3& JI&

Q-#,*-.*'88"01'(-&,*

JJ&

da&@(/(%3& T0"3,0:&=2$0#(,9& R2"#&_36#"#21&& _:"7(37& *R_E_-& b)0%#2/$23$2&c(62F:2&_:"7(37& !($#%/$%+9&*bc_!-&"31&/0+2#?#2/%)0F%3& :($#%/$%+9&*=^LaO&=^YE!O&'=a_!O& `<c!O&2,$;-& ^(:2?%6?b)(74,&& `%/(,#%3&L:(//(%3&^%:%7#"+49& *^Yb&`L^-&

!

=0+2#?#2/%)0F%3&!($#%/$%+9& ^(:2?#2/%)>21&E":"3&& =+2$,#%/$%+9& L)2$,#%3($/&&6%#& T0"3,0:&B%:+0F37&

4EM/*/0&G"$JB5-#-&*R7'G$%,*9=>>@J?@<*

G&HIJK&L1%"#1%&B4"#5%3& JH&

!(3(",0#(e"F%3& B%:+)2W(,9&

Q'#$,#F*

G&HIJK&L1%"#1%&B4"#5%3& JJ&

ISSCC 2015

M#5$%*#%$&K,*

• da&_B&
• =05?Pf3:&B!Y=&
• R28&!",2#(")/&&
• b)2W(5)2&

G&HIJK&L1%"#1%&B4"#5%3& JK&

="::"gO&<:(3("3O&B4"#5%3O&& R"3>2#O&_La!JJ&

!

=03O&_/4(4"#"O&B4"#5%3O&& S=^TLO&HIJK& ="::"gO&<:(3("3O& B4"#5%3O&R"3>2#&& LB=JK& !","&`">("O&M%)6O&B4"#5%3O&R==JK&

N5'#*'6-;#*8%-K;1#,V** $W$1#*-&*,-10$#:P*

G&HIJK&L1%"#1%&B4"#5%3& JP&

!

R&K;,#%0'"X4-&,;7$%*!11$8#'&1$*

JZ&

Alex Dommann EMPA

Pantelis Georgiou Imperial College London

Bio-ins io-inspir pired ed Semiconduct emiconductor

• rs for
• r

Healt Healthcar hcare e

Dr. . Pant antelis elis Geor Georgiou giou (pantelis@imperial.ac.uk) Centre for Bio-inspired Technology, Institute of Biomedical Engineering, Department of Electrical and Electronic Engineering Alt lter erna nativ ive e us uses es of

• f Silicon

ilicon Sympos mposium ium on

• n Emer

merging ging trends ends in in Elect lectronics

• nics

2nd

nd December

ecember 2014 2014

B23,#2&6%#&C(%?(3/+(#21&^2$43%)%79 &&&&&&&&&&&&&&&&&&&&&&&&& &&&&&&&a#;&`"3,2)(/&'2%#7(%0 & & & &&

Can we leverage on $1 trillion investment in Microelectronics to create more affordable healthcare?

CMOS Opportunities

• Scalable
• Repeatable
• Low cost
• Miniaturization
• System Integration
• Dedicated performance

(low noise, low power)

Page 21

21

Needs of Modern Healthcare

• Health services originally

designed to manage acute illness (i.e. infections and injury)

• Today, however 70-78% of

health budget expenditure is

• n Chronic Disease!

Personalised Healthcare

• Technology from Hospital to

the Home - New Wave Lifestyle

• Medical devices towards

consumer devices

B23,#2&6%#&C(%?(3/+(#21&^2$43%)%79 &&&&&&&&&&&&&&&&&&&&&&&&& &&&&&&&a#;&`"3,2)(/&'2%#7(%0 & & & &&

15 – 20 mins To read your DNA

• Min. 2 weeks

SNP Chip

DNA Testing Today DNA Testing at Imperial: No Lab

Page 22

22 Source: DNA electronics website (www.DNAe.co.uk)

Healthcare Application

• Point-of-Care Diagnostics
• Sequencing Technology
• f the Future

B23,#2&6%#&C(%?(3/+(#21&^2$43%)%79 &&&&&&&&&&&&&&&&&&&&&&&&& &&&&&&&a#;&`"3,2)(/&'2%#7(%0 & & & &&

Combined DNA Amplification and Detection in CMOS

C.Toumazou, et al.,“Simultaneous dna amplification and detection using a pH-sensing semiconductor system,” Nature methods, 2013.

23

Features

• Fully unmodified CMOS
• Detect Hydrogen ions using Ion-sensitive

Field Effect Transistors.

• Can use heating to amplify DNA through

PCR (polymerase chain reaction)

• Use a reference chamber to do differential

measurement and cancel out chemical drift.

Ion-Sensitive Field Effect Transistor

CMOS is cheap, integration is expensive, how do we break this barrier of co-design?

B23,#2&6%#&C(%?(3/+(#21&^2$43%)%79 &&&&&&&&&&&&&&&&&&&&&&&&& &&&&&&&a#;&`"3,2)(/&'2%#7(%0 & & & &&

Semiconductor DNA sequencing

Page 24

• Moores law has already found its way to DNA sequencing using ISFETs!

Source: ion Torrent website (www.iontorrent.com)

24

B23,#2&6%#&C(%?(3/+(#21&^2$43%)%79 &&&&&&&&&&&&&&&&&&&&&&&&& &&&&&&&a#;&`"3,2)(/&'2%#7(%0 & & & &&

Whole Genome Sequencing Costs

• Rapidly reducing costs!

Page 25

25

Wetterstrand KA. DNA Sequencing Costs: Data from the NHGRI Genome Sequencing Program (GSP)

The $100 Genome?

• When is Moore enough?
• Challenges:
• Things difficult to detect at

smaller volumes

• Noise, mismatch
• Reliability of sensors
• Micro-fluidic integration
• Data Bandwidth
• Bio-informatics, Big data

How can we scale CMOS sensors and fluidics reliably to exploit Moore?

B23,#2&6%#&C(%?(3/+(#21&^2$43%)%79 &&&&&&&&&&&&&&&&&&&&&&&&& &&&&&&&a#;&`"3,2)(/&'2%#7(%0 & & & &&

Diabetes

Insulin injection Glucose measurement

Sub-optimal treatment Current Treatment

• Insulin injection solves the problem

in the short term.

• Diabetics still spend a large amount
• f time in the hyper-glyceamia
• Leading cause of blindness, kidney

failure, heart disease, neurogenic disease.

Page 26

26

Artificial Pancreas

B23,#2&6%#&C(%?(3/+(#21&^2$43%)%79 &&&&&&&&&&&&&&&&&&&&&&&&& &&&&&&&a#;&`"3,2)(/&'2%#7(%0 & & & &&

Can we use a Bio-inspired Approach?

The Pancreas Semiconductor technology Novel Medical Devices Miniature, Low-power Biology Low-cost Microchips Therapy

Page 27

27

Pancreatic Beta Cells

B23,#2&6%#&C(%?(3/+(#21&^2$43%)%79 &&&&&&&&&&&&&&&&&&&&&&&&& &&&&&&&a#;&`"3,2)(/&'2%#7(%0 & & & &&

The Bio-inspired Artificial Pancreas

!*N$""1-7$*I%;,#*.;&K$K*0&0('()$*#-*1%$'#$*#5$*O-%"KY,*Z%,#*[0-JR&,80%$K* !%(Z10'"*B'&1%$',*.-%*6"--K*G";1-,$*4-&#%-"**

Bio-inspired, Physiological, On a Chip, with Low-power

Herrero P, Georgiou P, Oliver N, Johnston DG, Toumazou C, “A bio-inspired glucose controller based on pancreatic b-cell physiology”. J. Diabetes Sci.Technol.6,606–616. 2012

28

HI&/05X2$,&^#(")&

• &!2"3&%>2#3(74,&7)0$%/2&Z;K:!&
• &R%&49+%7)9$"2:("&

[DTT, 16(9): 550-557, 2014]

• Co-integration with wireless devices is currently difficult.
• Need to think about common wireless standards and

security.

• Needs a patient pull rather than an technological push!
• Design for Regulation!
• Need to guarantee safety! Built in self test, redundancy

B23,#2&6%#&C(%?(3/+(#21&^2$43%)%79 &&&&&&&&&&&&&&&&&&&&&&&&& &&&&&&&a#;&`"3,2)(/&'2%#7(%0 & & & &&

To conclude

Semiconductors for Healthcare

• Disruptive technology
• Leveraging on the $1 trillion investment over the

past 4 decades

• Realising medical systems like consumer devices
• Good for robust, low-cost, high density, low

power and high performance miniature systems Bio-Inspired Technology

• Implementing biology in modern technology to replace

biology

• Applications in chronic disease management
• Improved quality of life

“The future has already arrived. It's just not evenly distributed yet” William Gibson

Bruno Murari ST Microelectronics

Bruno Murari STMicroelectronics Scientific Advisor Montreux 2 dec. 2014

32

Inkjet cartridge

Sensors are Changing the World

Smart Me Fitness & Wellness

Help to lead healthier lives Optimize sports performance Early warning of illness

33

Smart Car

Reduce emissions Increase safety Save fuel

Smart City

Reduce traffic congestion Better use of resources Improve security

Smart Home

Make entertainment more interactive and immersive Increase comfort Save energy

Smart Me Healthcare

Empower patients Help physicians monitor and diagnose remotely

The Most Advanced MEMS Gyroscope

34

Drive mode

FCoriolis = -2m!z ! v

x y z v !z FCoriolis

Yaw mode Pitch mode Roll mode

AST Molecular Biology

Dynamic structural monitoring

Data processing, signal correlation

37

More than MEMS or Macro MEMS ? Surce:CSEM

38

Roland Thewes TU Berlin

Symposium on Emerging Trends in Electronics • Panel: Alternative Use of Silicon • Roland Thewes • Montreux, Switzerland • December 1 – 2, 2014 40

After the Gold Rush – Low Volume Biomedical CMOS Devices Creating Great Value?

Roland Thewes TU Berlin, Berlin, Germany roland.thewes@tu-berlin.de

Montreux, Switzerland December 1 - 2, 2014

Symposium on Emerging Trends in Electronics Panel: Alternative Use of Silicon

Symposium on Emerging Trends in Electronics • Panel: Alternative Use of Silicon • Roland Thewes • Montreux, Switzerland • December 1 – 2, 2014 41

Semiconductor Industry in the Year 2000:

The “Next Big Thing”?

Chip Market [Billion US$] Growth rate with respect to former year [%]

Source: WSTS

Year 2000:

• 200 billion US$ chip market frontier

exceeded for the first time

• “.com bubble”
• Huge growth of the amount of

biotech companies and startups

• Business developers in

semiconductor industry speculated that bio could be “the next big thing” also for the chip industry: time maturity

Bio ! year 2000

Symposium on Emerging Trends in Electronics • Panel: Alternative Use of Silicon • Roland Thewes • Montreux, Switzerland • December 1 – 2, 2014 42

• Today:

Total medical semiconductor revenue > 4 billion US$.

• However, most important segments

are:

• Home care including health and

wellness applications (blood pressure, heart rate, and glucose monitoring, …)

• Clinical including medical imaging

(portable medical electronics as portable ultrasound devices, portable ECG devices, …)

• This includes many “standard ICs” – possibly somewhat optimized for biomedical

purposes – such as processors and memory.

• So, where and how do customized CMOS biochips contribute?

Sources: WSTS, IC insight

Semiconductor Industry in the Year 2014:

The “Next Big Thing”?

Symposium on Emerging Trends in Electronics • Panel: Alternative Use of Silicon • Roland Thewes • Montreux, Switzerland • December 1 – 2, 2014 43

CMOS Devices for Biomedical Purposes

Devices and Applications

today’s volume (?) product emerging devices / fields under development

Pace maker, Zarlink Hearing aid, Phonak Blood glucose meter, Bayer Ultrasound machine, portableultrasounds Wearable EEG headset, IMEC Deep brain stimulator, Medtronic (fig. right from www.nature.com) Utah BMI Cochlear implant (from K. Wise, IEDM 2002) DNA microarray, Siemens DNA sequencing chip, Ion Torrent Retinal implant (from

• M. Ortmanns, JSSC 2007)

Next generation DBS and BMI devices, Neuronex “Neurochips”, Max-Planck Society

Symposium on Emerging Trends in Electronics • Panel: Alternative Use of Silicon • Roland Thewes • Montreux, Switzerland • December 1 – 2, 2014 44

Where does Innovation Come from?

Two Established Applications …

• Deep Brain Stimulation
• The device implanted in the brain stimulates the brain

and is controlled by the device implanted underneath the skin within the thorax region.

• Approved for human therapies since 2009.
• Closed loop systems under development
• Volume:

(S. Oesterle, Medtronic, Plenary Talk ISSCC 2011): “… worldwide 1 device every 30 minutes …”

www.nature.com

• Cochlear Implants
• restoring hearing to the profoundly deaf by means of

auditory nerve stimulation

• Internal devices: array of electrodes wound through

the cochlea to stimulate the auditory nerve, receiver, stimulation circuitry. External devices: microphone(s), speech processor, power / data transmitter

• Development since more than 30 years
• Volume: 300,000 worldwide on December 31, 2011.

http://commons.wikimedia.org/wiki/ File:Blausen_0244_CochlearImplant_01.png

Symposium on Emerging Trends in Electronics • Panel: Alternative Use of Silicon • Roland Thewes • Montreux, Switzerland • December 1 – 2, 2014 45

What Do CMOS Chips Contribute?

Consider CMOS Chip Based DNA Sequencing …

• DNA sequencing: determination of the
• rder of the four DNA bases (A, T, C, G)
• Human DNA: 3.2 109 base pairs
• Market estimation for 2015: 2 billion US$
• Next generation sequencing (NGS):

faster + cheaper

• Ion-Torrent: CMOS chip with ISFET

array, “sequencing-by-synthesis”

• What else is left to do?
• 1. Purification of Genomic DNA from cells (lysis, breaking the cells’ nuclei, …)
• 2. Fragmentation (random process: fragments must “overlap”, length up to 1 kb)
• 3. Amplification (fragments serve as “vectors”, vectors are cloned, …)
• 4. Determination of sequence of bases, i.e.: fragment-based sequencing (CMOS!)
• 5. Determination of the entire strand sequence by overlapping fragment data
• J. M. Rothberg et al, Nature 475, 348–352 (21 July 2011)

Symposium on Emerging Trends in Electronics • Panel: Alternative Use of Silicon • Roland Thewes • Montreux, Switzerland • December 1 – 2, 2014 46

Biomedical CMOS Chip Properties

… Considered from the User’s Standpoint

• Users and customers (who are not electronic engineers) do not care what is

inside the chip

• However, they appreciate functionality which is not available using other tools or

technologies CMOS technology, appropriate and innovative circuit design, assembly techniques, software, … are enablers!

• Example “Neurochips” with high spatiotemporal resolution:
• first publications roughly 10 years ago
• commercialization: started recently and is on-going
• discussed volumes: few 1000 devices per year

Symposium on Emerging Trends in Electronics • Panel: Alternative Use of Silicon • Roland Thewes • Montreux, Switzerland • December 1 – 2, 2014 47

Summary

Low Volume Biomedical CMOS Chips

• Create value in an ethical, societal, and commercial sense although

related applications are niches.

• Related business models must be identified to make technical
• pportunities successful in real-life applications.
• As engineers we need to try to understand the entire application chain,

and not to develop solutions and then try to identify the problem.

• Scaling helps in that sense that older technologies get affordable to

create a business case. Many customized biomedical chips rely on technologies with > 100 nm feature size.

• Electronics can be an inevitable enabler (but only one piece of the entire

puzzle).

Alternative Use of Silicon

Edoardo Charbon TU Delft Alex Dommann EMPA Pantelis Georgiou Imperial College London Bruno Murari ST Microelectronics Roland Thewes TU Berlin Chair: Giovanni De Micheli, EPFL