Electrolysis Without Membranes

Electrolysis without Membranes

Glen O’Neil, Oyin Talabi, David Brown, Cory Christian, Ji Qi, Jack Davis, Anna Dorfi Dan Esposito Department of Chemical Engineering Lenfest Center for Sustainable Energy Columbia University

Closing the Carbon Cycle Conference Tempe, AZ, September 29th, 2016

9/29/16 D. Esposito, Closing the Carbon Cycle 1 Mission The mission of the Lenfest Center for Sustainable Energy (LCSE) is to advance science and develop innovative technologies that provide sustainable energy for all humanity while maintaining the stability of the Earth’s natural systems.

Research Themes Six interconnected, topical research areas that fall under the overall theme of sustainable energy conversion and utilization pathways:

I. Novel Materials/Nanotechnology for energy conversion, utilization and storage with a reduced environmental footprint.

II. Novel Reaction Pathways for sustainable energy materials conversion throughout the engineered and natural elemental cycles, including innovative device development (e.g. 3-D printed reactor systems).

III. Catalysis for novel reaction pathways.

IV. Separations using smart, multi-functional materials for sustainable energy and materials.

V. Energy Storage and Systems Integration for optimized deployment of renewable energy.

VI. Earth Systems for sustainable energy extraction, conversion and waste storage (e.g. waterless or CO2-rich fracking of shale, enhanced oil recovery, geothermal heat recovery with integrated CO2 conversion, and CO2 storage).

9/29/16 D. Esposito, Closing the Carbon Cycle 2 Electrochemical Production of Fuels

(+) (-) H2O Electrolysis V - Red.: + - e 2H +2e → H2 - Ox.: - + e H2O →2e 2H +0.5O2

Overall: H2O → H2 +0.5O2

O2 H+ H2

H2O

Anode Cathode membrane Side-view of PEM electrolysis cell.

9/29/16 D. Esposito, Closing the Carbon Cycle 3 Electrochemical Production of Fuels

(+) (-) H2O Electrolysis V - Red.: + - e 2H +2e → H2 - Ox.: - + e H2O →2e 2H +0.5O2

Overall: H2O → H2 +0.5O2 C H O O2 x y z H+ CO2 Electrolysis

+ - Red.: xCO2 + 2nH + 2ne →CxHyOz H O CO2 - + 2 Ox.: n(H2O →2e + 2H + 0.5O2) Overall: xCO + nH O → C H O Anode Cathode 2 2 x y z membrane Side-view of PEM electrolysis cell. CxHyOz= CO, CH4, HCOOH,

CH3OH, C2H4, and more

9/29/16 D. Esposito, Closing the Carbon Cycle 4 Sometimes it may seem as if we have two choices..

“Hydrogen Renewable Economy” “Hydrocarbon Economy” H2 CxHyOz

http://blog.capterra.com/wp-content/

9/29/16 D. Esposito, Closing the Carbon Cycle 5 “Hydrogen-carbon Economy”

“Hydrogen Renewable Economy” “Hydrocarbon Economy” H2 CxHyOz But it is important to

remember that H2 and CHO’s are complementary http://blog.capterra.com/wp-content/

9/29/16 D. Esposito, Closing the Carbon Cycle 6 Examples of the “Hydrogen Carbon” Economy Example 1: Fischer Tropsch Process

Electrolyzer CO2 CO + O2 liquid

Fischer- (-CH)n H2 Tropsch H2O

Example 2: Sabatier Process Example 3: Methanol Synthesis

3a. CO + 2H2 → CH3OH CO2 + 4 H2 → CH4 + 2 H2O

3b. CO2 + 3H2 → CH3OH + H2O

Ethylene, Propylene, acetic acid, and others

9/29/16 D. Esposito, Closing the Carbon Cycle 7 Economics of H2 from Low Temperature Water Electrolysis What’s it going to take?!

$2-4 /kg H2 DOE Hydrogen & Fuel Cells Program target cost for production & delivery in order to compete 1. Polymer Electrolyte membrane with gasoline at the pump.[1] (PEM) electrolyzer http://www.hydrogenics.com (+) (- ) Where are we today? O2 H2 Gas collection [1] $4-5 /kg H2 30% KOH or NaOH (Production Cost Only)

Separation Diaphragm Note: Production of H2 from CH4- [2] 2. Alkaline electrolyzer (unipolar) reforming is ≈$2/kg H2. (2012) Harrison, Levine, “Electrolysis of Water” (2007)

[1.] 2012 DOE‐FCTP MYRD&D cost status and targets for Hydrogen Production. [2] S.9/29/16 Dillich , et D. al., Esposito, “Hydrogen Closing Production the CostCarbon Using Cycle Low-Cost Natural Gas”, DOE report, (2012), available online. 8 Economics of H2 from Low Temperature Water Electrolysis

Breakdown of H2 production costs by water electrolysis. Data from [1] Soft Costs^

Capital 1. Polymer Electrolyte membrane (PEM) electrolyzer Costs http://www.hydrogenics.com Electricity (+) (- ) O2 H2 [1] Study by Strategic Analysis based on $0.07/kWh Gas collection electricity & PEM electrolyzer system(≈$100/kW) with 97% availability factor.

30% KOH or NaOH ^Includes indirect, O&M, and replacement

Separation The cost of H2 produced by electrolysis is Diaphragm usually dominated by the cost of electricity. 2. Alkaline electrolyzer (unipolar) Harrison, Levine, “Electrolysis of Water” (2007)

[1.]9/29/16 W. Colella D., et Esposito, al., “Techno Closing-economic the Analysis Carbon of PEMCycle Electrolysis for Hydrogen Production”, (2014) available online.9 Considerations for H2 Production from PV/Wind-Electrolysis: 1. The price of electricity from renewables is decreasing……….

2.4 cents/kWh

http://www.pv-magazine.com/news/details/beitrag/breaking--world-record-low-price-entered-for- solar-plant-in-abu-dhabi_100026145/#axzz4LYZnCkaP

9/29/16 D. Esposito, Closing the Carbon Cycle 10 Considerations for H2 Production from PV/Wind-Electrolysis: 1. The price of electricity from renewables is decreasing……….

Beyond

DOE target (Upper limit)

Influence of cost of electricity on cost of H2. Analysis based on electrolyzer efficiency of 75% (HHV). Note: DOE target is upper limit for production and delivery.

Beyond Sunshot: http://energy.gov/eere/sunshot/photovoltaics-research-and-development 9/29/16 D. Esposito, Closing the Carbon Cycle 11 Considerations for H2 Production from PV/Wind-Electrolysis: 2. BUT often only for 25-40% of the time………. Time of use pricing

Average capacity factor for utility- scale PV power generators in the U.S. (2015): [1] Time of use(TOU) pricing scheme in CF= 28.6%. Ontario Canada (Summer, Weekdays)[2] Low PV/Wind capacity factor and time of use pricing mean that electrolyzers would likely sit idle most of the time.

[1.] U.S. EIA, Electric Power Monthly, (2015) https://www.eia.gov/electricity/monthly/epm_table_grapher.cfm?t=epmt_6_07_b [2.]9/29/16 “Time of UseD. Esposito,Pricing in Ontario” Closing https ://www.keyframe5.com/smartthe Carbon Cycle -meters-time-of-use-tou-in-ontario/ 12 Cost Considerations for H2 Production from PV/Wind-Electrolysis: Capacity factor for PV/Wind Electrolysis or time-of-use pricing

Key Assumptions Electrolyzer • 4 ₵/kWh electricity • Electrolyzer efficiency of 75% (HHV) • System lifetime=10 yrs • non-discounted analysis • Installation=12% CapEx

Electrolyzer capacity factor Relationships between capital cost and capacity factor at fixed cost of electricity. Capital costs become much more important at low capacity factors.

9/29/16 D. Esposito, Closing the Carbon Cycle 13 Capital Costs for a PEM Electrolyzer System PEM Electrolyzer System Capital Cost Breakdown [1] (+) (-) V e- e-

O2 + balance H H of 2 system Stacks^ H2O

Anode Cathode membrane Side-view of PEM electrolysis cell.

^ ≈ 60% of stack cost is from MEA (membrane + electrodes)

•Balance of system (BOS) components are important •Electrolyzer stack is the most expensive single component

[1.] W. Colella, B. James, et al., “Techno-economic Analysis of PEM Electrolysis for Hydrogen Production”, (2014) available at: http://9/29/16energy.gov/sites/prod/files/2014/08/f18/fcto_2014_electrolytic_h2_wkshp_colella1.pdf D. Esposito, Closing the Carbon Cycle 14 PEM Electrolyzer Stack

Single cell

PEM Stack (6 cells). [1] PEM electrolyzer http://www.hydrogenics.com

•60% of stack cost is from MEA[2] •An MEA design requires many parts and assembly steps[1]

[1.] V. Mehta, J. Power Sources, 114 (2003). [2] W9/29/16. Colella ,D. et Esposito,al., “Techno Closing-economic the Analysis Carbon of PEM Cycle Electrolysis for Hydrogen Production”, (2014) available online.15 PEM Electrolyzer Stack

Key components in PEM Stack • End plates • Flow field plates • Gaskets/seals • Fasteners

• Membrane • Gas diffusion layer • Anode catalyst • Cathode catalyst PEM electrolyzer Membrane electrode assembly (MEA) http://www.hydrogenics.com

•60% of stack cost is from MEA[2] •An MEA design requires many parts and assembly steps[1]

Key components in PEM Stack • End plates Question: At the most • Flow field plates • Gaskets/seals basic level, how many • Fasteners parts are really needed? • Membrane • Gas diffusion layer • Anode catalyst • Cathode catalyst • Device body

•60% of stack cost is from MEA[2] •An MEA design requires many parts and assembly steps[1]

H2O

membraneless microfluidic device

Co-laminar membraneless fuel cells,[6,8] and flow batteries[7,8] Membraneless electrolyzer with separation based on the Segre-Silberberg effect.[9] Advantages of a Membraneless Cells: • Decreased capital costs Membraneless Laminar flow Cells: • Electrolyte-agnostic • Utilize “flow-by” electrodes with • No membrane degradation and fouling inherent limitations in scalability[8] • Simple: new manufacturing possibilities

[6.] E. Choban, P. Kenis, et al., J. Power Sources 128 (2004) [8.] E. Kjeang, et al., J. Power Sources, 260 (2014) [9.] S. Hashemi, D. Psaltis, et al., Energy Env. Sci., (2015). [7.] Braff,9/29/16 Bazant D., etEsposito, al., Nat. Comm. Closing (2013). the Carbon Cycle 18 Membraneless Electrolyzer w/ Flow-Through Electrodes[1],[2] Low fluid flow rate Zoomed-in View H2O H2O

O2 H2

O2 H2 High fluid flow rate

Mesh Mesh

Anode Cathode O2 H2 H O H2O 2 Membraneless electrolyzer based on Illustration of “void fracture” caused circular mesh electrodes in a face-to- by gas accumulation in the absence face configuration.[1] of insufficient fluid flow.[1]

[1.] M.I. Gillispie, et al, J. Power Sources, 293 (2015). [29/29/16.] J. Hartvigsen D. Esposito,, et al., ECS Closing Transactions, the Carbon 68 (2015) Cycle 19 Membraneless Electrolyzer with Angled Flow-Through Electrodes

+ - 2H +2e → H2 Electrolyte + electrolysis products Aqueous (H2 or O2) electrolyte

Mesh flow-through electrode + - H2O → 0.5O2+2H +2e Schematic top-view of membraneless electrolyzer Advantages of Proposed Design: based on flow-through mesh electrodes[1,2] • Same advantages over conventional PEM devices as other membraneless devices • Inherently more scalable • Expected to be more tolerant to flow-characteristics than co-laminar flow cells • Extremely simple design should decrease assembly/ manufacturing costs

[1.] D. Esposito & G.D. O’Neil, U.S. Patent application, (2015). [2.] O’Neil,9/29/16 Christian, D. Esposito, Brown, Esposito,Closing J.the Electrochem Carbon Cycle. Soc. 163 (11) F3012-F3019 (2016) 20 Flow-Through Electrodes: Electrodeposited Pt on Ti Mesh

1 mm

Titanium (Ti) mesh flow-through electrode SEM image of Pt deposits with electrodeposited Pt catalyst

Pt-Ti mesh anodes and cathodes fabricated by electrodeposition

9/29/16 D. Esposito, Closing the Carbon Cycle 21 Electrolyzer Fabrication with 3D Printing

Baffle Divider H2

20 min. Cathode (- ) O2 Flow channel

60 min. Anode (+)

Channel dimensions Width: 1.3 cm Electrolyte Height: 0.7 cm Inlet Length: 7.0 cm 150 min.

9/29/16 D. Esposito, Closing the Carbon Cycle 22 Does it Work?: Basic Operation of Angled Flow Through Electrodes

5 mm Conditions (+) H2O (-) • Flow rate: 6.5 mL/s (Re≈35) • 45⁰ cell

• Electrolyte: 0.5 M H2SO4 • Current density=100 mA/cm2

High speed video of gas generation and removal from 45 degree cell. Video was recorded at 500 frames per sec. and converted to “edge detection” view.

9/29/16 D. Esposito, Closing the Carbon Cycle 23

Does it Work?: Basic Operation of Angled Flow Through Electrodes

2 Flowing electrolyte - 30 degree cell in 0.5 M H SO (-) 2 4

Flow

(+) Stagnant

(-)

urrent density / mA cm mA / density urrent C

Electrolysis current at 2.5 V in 0.5 M H2SO4 with (+) various fluid velocities.

Fluid flow sweeps bubbles away from electrodes into downstream collection channels and significantly reduces mass-transport and iR losses associated with bubbles on the electrode surface

9/29/16 D. Esposito, Closing the Carbon Cycle 24 Electrolysis Efficiency: =η= ΔG/nF = ΔEcell efficiency ΔV ΔV

30⁰ cell 30⁰ cell 0.5 M H2SO4 0.5 M H2SO4

η=56% η=51% ΔV

ΔEcell

2-electrode current-voltage curve Electrolysis efficiency based on

taken in flowing electrolyte. ΔG=-237 kJ/mole H2.

• Electrolysis efficiency in 0.5 H2SO4 is OK but relatively low compared to commercial electrolyzers.

9/29/16 D. Esposito, Closing the Carbon Cycle 25 Evaluating Electrolyzer Performance

ΔG=237 kJ/mole Electrolysis efficiency: η= ΔG/nF = ΔEcell ΔE o=1.23 V ΔV ΔV cell Applied =ΔV = ΔE + η + η + η + iR Voltage cell HER OER mt s

Cell Potential Kinetic Mass Ohmic (Nernst Eqn.) overpotential transfer losses losses losses

• Losses in electrolyzer can be broken down into separate components as commonly performed in fuel cell loss-analyses

9/29/16 D. Esposito, Closing the Carbon Cycle 26 Loss Analysis for Operation in 0.5 M H2SO4

+ - H2O → 0.5O2+2H +2e ηMT < 0.05 V

iRS=0.19 V ηOER

ηHER =0.18 V ΔV=2.4 V

@100 ηOER =0.91 V -2 + - mA cm 2H +2e → H2

ΔEcell=1.23 V ηHER

Voltage losses for 30⁰ cell at 100 mA cm-2 in

0.5 M H2SO4 under flowing electrolyte. 3-electrode CV measurement for Pt-Ti mesh electrode in 0.5 M H2SO4.

• Oxygen evolution kinetic overpotential losses greatly limit performance • Future efforts to move away from Pt

9/29/16 D. Esposito, Closing the Carbon Cycle 27 Electrolyzer Operation in Different Electrolytes

Efficiency 2 - @ 50 mA cm-2 based based Electrolyte on ΔG on HHV

1 M Na2SO4 41 % 50 %

100 mA cm-2 0.5 M H2SO4 56 % 67 % 1 M KOH 67 % 81 % 50 mA cm-2

^ evaluated at an operating current density of 50 mA/cm2 under flowing

Current density / mA cm density mA / Current electrolyte

• Device operation in three different solutions highlights versatility of membraneless electrolyzers. • Highest efficiency observed for KOH electrolyte (best kinetics)

[1.]9/29/16 O’Neil, Christian, D. Esposito, Brown, Closing Esposito, the J. Electrochem Carbon Cycle. Soc. 163 (11) F3012-F3019 (2016) 28 Collection Efficiency and H2 Purity

Conditions: -30⁰ electrolyzer -215 mA cm-2

-Electrolyte: 0.5 M H2SO4 -Flow velocity: 6.6 cm s-1

Collection efficiency:

H2 collected ηc = H2 generated

Volume of H2 collected in cathode effluent channel compared to 100% collection efficiency as expected from Faraday’s Law.

• Collection efficiencies of ≈ 90 % demonstrated • GC analysis of collected gas gives ≈ 2.8% cross-over rate

9/29/16 D. Esposito, Closing the Carbon Cycle 29 Can the electrolyzer be simplified even further?

Side view of passive membraneless electrolyzer assembly consisting of Electrolysis cell for evaluating the “asymmetric” electrodes. performance of asymmetric electrodes.[1]

[19/29/16.] J. Davis, D. J. Qi,Esposito, D. Esposito, Closing (In Preparation) the Carbon . Cycle 30 Can the electrolyzer be simplified even further?

(-) (+)

Sensor

Dissolved H2 detected in between two mesh electrodes as a function of current density using an High speed video of asymmetric Pt/Ti electrodes electrochemical sensor. -2 [1] operating at 40 mA cm in 0.5 M H2SO4.

[19/29/16.] J. Davis, D. J. Qi,Esposito, D. Esposito, Closing (In Preparation) the Carbon . Cycle 31 Conclusions 1. Bringing down the capital cost of electrolyzers will become increasingly important for making fuels produced from renewable energy competitive with fossil fuel energy.

2. A simple, “3-component” membraneless electrolyzer was made and demonstrated in this study.

3. Much room for improvement from initial results, but potential exists for simple, low- cost electrolyzers beyond the MEA cell architecture.

9/29/16 D. Esposito, Closing the Carbon Cycle 32 Acknowledgements • Esposito Research Group at Columbia University -Glen O’Neil -Ji Qi -Cory Christian -Jack Davis -David Brown -Oyin Talabi -Anna Dorfi

• Columbia Start-up funds • Chris Hawxhurst (GC)

9/29/16 D. Esposito, Closing the Carbon Cycle 33 Questions?

[email protected] https://danesposito.wordpress.com/9/29/16 D. Esposito, Closing the Carbon Cycle 34