Highly Selective and Stable Multivariable Gas Sensors...• Several independent responses from individual sensor • Disruptively overcome insufficient selectivity of existing sensors

R.A. Potyrailo 2017 1Public Copyright © 2017 General Electric Company - All rights reserved

18th Annual SOFC Project Review Meeting, Pittsburgh, PA, June 12-14, 2017

Highly Selective and Stable Multivariable Gas Sensors

for Enhanced Robustness and Reliability of SOFC Operation

GE:

Radislav Potyrailo, Vijay Srivastava, Brian Scherer,

Majid Nayeri, Andrew Shapiro

SUNY Poly:

Michael Carpenter, Nicholas Karker

NETL Contract FE0027918


Background

• needs to improve cost-effectiveness

• enhance operation reliability

• early diagnostics of potential upsets

• ability to operate cells at optimum conditions

Goals for wide adoption of SOFC

Technical strategy

Phase 1 activities:

• develop sensing materials

• perform lab tests and field validation

Phase 1 outcomes:

• fundamental understanding of multivariable gas sensors at high temperatures

• enable cost-effective and stable sensors for SOFC applications

Project objective

Fuel Processing

Unit

In situSensor

SOFCfor

field validation

Technical approach

• implement new generation of gas sensors (multivariable sensors)

• leverage design rules of multivariable sensors for SOFC monitoring

• dynamic information about SOFC reforming gases

gases in Phase 2: 15-20% H2, 10-20% CO, 5-20% CO2, 2-15% CH4, 40-60% H2O

gases in Phase 1 are subset from Phase 2


Proposed sensor vs available offerings

Sampling

Cost

Available$5-10 sensor, $50 – 10,000 systempoor selectivity, in situ operation at ambient temperature

Available$250,000 system,

high selectivity, extensive sampling,

lab operation

In situ sensor

Proposed$10,000 - 20,000 system, high selectivity, in situ operationat high temperature

In Situ

Sensor requirements

• High reliability (accuracy, stability)

• Low initial / operation cost

• Low power consumption


Commodity gas detection:Serving sensing needs in established markets

• Mature status-quo technology

• Widely available

• Interchangeable

• Inexpensive

Photo by Potyrailo


Commodity gas detection: Challenges outside established markets

“The biggest headaches are caused by interfering chemicals…”Lewis, Edwards, Nature, 2016, 535, 29-31

Status quo of conventional sensors

Existing sensors do not meet monitoring demands

of new growing markets

0

100

200

300

400

500

600

Eth

yle

ne

oxid

e

Carb

on

mo

no

xid

e

Eth

an

ol

Me

tha

no

l

Iso

pro

pa

no

l

Iso

bu

tyle

ne

Bu

tad

ien

e

Eth

yle

ne

Pro

pe

ne

Vin

yl ch

lori

de

Vin

yl a

ce

tate

Fo

rmic

acid

Eth

yl e

the

r

Fo

rma

ldeh

yd

e

Ele

ctr

och

em

ica

l se

nso

r:

rela

tive g

as s

en

sitiv

ity (

%)

Electrochemical Non-dispersive IR

0

100

200

300

400

500

600

700

Meth

ane

Eth

ane

Pro

pane

n-B

uta

ne

n-P

enta

ne

n-H

exane

Non-d

ispers

ive infr

are

d s

ensor:

rela

tive g

as s

ensitiv

ity

(%)


Sensor arrays as accepted compromise

Sensor arrays:

Compromise between selectivity of sensor system and system complexity

Stability of

sensor arrays

0 101.0

1.8

Drift of each sensor (%)

Norm

aliz

ed

Pre

dic

tion

Err

or 20-sensor

array


Modern approaches for detection of volatiles with high selectivity:

Solving demanding measurement needs with available tools

Cooks, Purdue University Pine-environmental.com

Portable MS Portable GC

Performance of competing approaches

Existing approaches for high value gas detection

•Environmental

•Homeland Security

•Industrial Hygiene

•Petroleum / Biofuel

•Chemical

•Agriculture

•Food & Beverage

•Pharmaceutical

•Forensic

Multi-analyzer


11

Breaking status quo:

multivariable photonic gas sensors

Individual multivariable sensors:

• Several independent responses from individual sensor

• Disruptively overcome insufficient selectivity of existing sensors

Potyrailo et al.Nature Photonics 2007

Nature Photonics 2012

Proc. Natl. Acad. Sci. USA 2013

Nature Communications 2015

Chemical Reviews 2016

Carpenter et al.Anal. Chem. 2012

Beilstein J. Nanotechn. 2012

ACS Nano 2014

Nanoscale 2015


Physics of color formation:Value for multivariable photonic sensors

Combined effects of multilayer interference and diffraction

produce iridescence in natural Morpho butterfly scales

Potyrailo et al. Nature Photonics 2007

Potyrailo et al., Proc. Natl. Acad. Sci. U.S.A. 2013

Potyrailo et al. Nature Communications 2015


Bio-inspired gas sensors

Natural Bio-inspired

2 mm


Potyrailo et al., Proc. Natl. Acad. Sci. U.S.A. 2013

Potyrailo et al. Nature Communications 2015

Data analytics

Multivariableresonant sensor

Sensor node

System design

System concept

•Spatial orientation of surface functionalization

•Chemistry of surface functionalization

•Extinction and scattering of nanostructure

Design rules for gas-selectivity

control


Previous results for multivariable optical sensors

Factor 1 300

0

-12

Blank

CO

H2

Fa

cto

r 2

Polar

Non-polar

Ch

em

ica

l p

ola

rity

gra

die

nt

Vapors

in gas phase

Adsorbed vapors

NP-

NP-

P-

NP-

NP-

NP-

P-

P-

P-


Unusually high vapor response selectivity of natural Morpho scales

Potyrailo et al. Proc. Natl. Acad. Sci. USA 2013

Understanding of origin of selective vapor response

Ben = benzene

ACN = acetonitrile

MEK = methyl ethyl

ketone

MeOH= methanol

H2O = water

Carpenter et al. Anal. Chem. 2012

Potyrailo et al.

Nature Communications

2015

High temperature sensing materials with diverse gas response mechanisms

Design rules for bio-inspired highly selective gas sensors


1 mm 1 mm

1 mm2 mm

5 mm

Diversity of fabricated bio-inspired nanostructures

R. Potyrailo, Chem. Rev. 2016

Selective etching of ALD material

FIB CVD Electron-beam lithography

FIB CVD = Focused ion-beam

CVD = chemical vapor deposition

ALD = atomic layer deposition

CVD, UV litho, chemical etching

Conventional photolithography and chemical etchinglaser interference litho

Double-molding


Examples of 3D sensing materials

utilized in this program

Innovative design of sensing structures for high selectivity gas detection


Sample Holder

CCD Spectrograph

Lamp

8-channel gas-delivery system

Gas cabinets

In-house built multi-channel vapor generation and mixing systems

• Computer control

• Flexible flow profiles

• Multi-gas mixing

Spectral imaging systemSpectroscopy system


0Time (h)

10 0Time (h)

12 0 2 4 6 8 10

693.4

693.6

693.8

694.0

694.2

694.4

694.6

694.8

695.0

695.2

695.4

Se

nso

r re

sp

on

se

(a

rb

u

nits)

Time (hours)

0 10Time (h)

Sensor

Respo

nse (

Arb

. U

nits)

Sensor

Respo

nse (

Arb

. U

nits)

Sensor

Respo

nse (

Arb

. U

nits)5% 8% 10%

10% 15% 20%

10% 20%15%2

Example of single-gas responses

Response to H2 Response to CO2Response to CO

On track for needed response sensitivity and speed


Tools for data analysis of multivariable sensors:machine learning, data analytics, multivariate statistics

Wikipedia.org

Increasing role of data analytics in high performance sensing


Toward selective and stable bio-inspired gas sensors for solid oxide fuel cells

Stability of multivariable sensorField test

Data analytics

Multivariableresonant sensor

Sensor node

System design

System concept


Initial response stability Example: H2 gas

Sensor

response (

arb

units)

Time (days)

Baseline studies of sensor stability


Short-term responses of sensor:

analyte, interference, their mixtures

Analyte

1 2 3 4Interference

1 2 Interference

1

Analyte

1 2 3 4

Analyte

1 2 3 4

Interference

2

Ga

s flo

ws

Re

fle

cta

nce

Re

fle

cta

nce

Time 0 250

Wavelength WL #1

Wavelength WL #2

Excellent short term response for selective analyte quantitation


Long-term responses of sensing nanostructure:

analyte, interference, their mixtures

WL = wavelengthBS = baseline

Re

fle

cta

nce

Re

fle

cta

nce

Time 0 3000

WL #1

WL #2

BS

BS

Long term response exhibits strong baseline drift


Corrected long term sensor baseline drift

Predicted analyte concentrations

from long-term responses

Time 0 3000

Conventional method

New method

Pre

dic

ted

co

nc.

Pre

dic

ted

co

nc.


Fuel Processing

Unit

In situSensor

SOFCfor

field validation

Current work Upcoming work

Summary

•Design of bio-inspired sensing materials for simultaneous quantitation of several gases

•Refined material-design rules to operate at elevated temperatures

•Advancing fundamental understanding of multivariable gas sensors at high temperatures

• Initial stability tests started


Acknowledgements

• GE Fuel Cells SOFC Team

• GE Global Research Team

• SUNY Albany team

• Steven Markovich @ DOE/NETL

• Funding provided by the US Department of Energy through cooperative agreement FE0027918

This material is based upon work supported by the Department of Energy under

Award Number FE0027918. However, any opinions, findings, conclusions, or

recommendations expressed herein are those of the authors and do not

necessarily reflect the views of the DOE.

Documents

Highly Selective and Stable Multivariable Gas Sensors...• Several independent responses from individual sensor • Disruptively overcome insufficient selectivity of existing sensors