Electrocatalysis with Discrete Transition Metal Complexes in Energy Conversion Systems Chris Chidsey Department of Chemistry, Stanford University

Transcription

1 Electrocatalysis with Discrete Transition Metal Complexes in Energy Conversion Systems Chris Chidsey Department of Chemistry, Stanford University ther Team Members Prof. Dan Stack Prof. Bob Waymouth ick Conley Charles McCrory on Painter Xavier ttenwaelder Anando Devadoss David Pearson Liezel Labios GCEP Symposium, September 19, 2006

2 utline ydrocarbon Fuel Cells: The ptimal Mobile Power Source? Electrocatalytic eversibility: From Efficient Power Delivery To Efficient Energy Storage Molecular Electrocatalysts on Carbon Electrodes The Ubiquitous Problem of 2 eduction The Many Challenges of ydrocarbon xidation C- Activation, C-C Activation, C xidation

3 Polymer-Electrolyte-Membrane Fuel Cell e motor - e - oxidation catalyst fuel graphitic carbon electrode reduction catalyst oxygen graphitic carbon electrode polymer electrolyte membrane (PEM)

4 The Atkins Diet for Fuel Cells: Benefits of Long-Chain ydrocarbons for Mobile Fuel Cells Twice the volumetric energy density of methanol kerosene or diesel ~ 8 kwh/l methanol ~ 4 kwh/l Twice the gravimetric energy density of methanol kerosene or diesel ~ 10 kwh/kg methanol ~ 5 kwh/kg Water insoluble, low flammability, low toxicity Existing production and delivery infrastructure educed leakage through polymer-electrolyte membrane

5 Electrocatalyst Efficiency Fuel best existing hydrogen 2 free energy wasted as heat of fuel oxidation heat free energy converted to work work free energy wasted as heat of 2 reduction heat xidant best existing oxygen 2 best existing methanol C 3 heat work heat best existing oxygen 2 hydrocarbon (C 2 ) n ow much work can be obtained and heat avoided with discrete electrocatalysts? oxygen 2 reversible potentials Electrochemical Potential (V vs. / 2 ) E 0 / 2 E 0C22/C2 E 0 2 / 2

6 Can we find catalysts to insert into existing device? The basic unit cell of a proton exchange membrane fuel cell is assembled in the required numbers to deliver the necessary power output. The component layers of the membrane electrode assembly (two catalysed substrates and the PEM electrolyte) are laminated together and located between the flow field plates to form the unit cell (T.. alph and M. P. ogarth, Platinum Metals ev., 2002, 46, (1), 3 14.)

7 Leverage ngoing Engineering of Polymer Electrolyte Membrane (PEM) Fuel Cells Benefits: ighly engineered mass and ion transport obust, high-surface-area carbon-particle electrodes ighly developed electrical interconnects (bipolar plates, corrosion resistance, etc.) ighly developed fuel and air systems Well developed thermal and water management Challenge: Efficient and stable electrocatalysts in strong acid

8 Typical ydrogen Fuel Cell Losses Schematic showing the typical cell potential versus current density output from an MEA operating on pure hydrogen and pure oxygen. The major factors that control the MEA performance in the various regions of the cell potential versus current density curve are identified along with their relative contribution to the electrical efficiency losses (T.. alph and M. P. ogarth, Platinum Metals ev., 2002, 46, (1), 3 14.)

9 Toward Electrocatalytic eversibility For Efficient Power Delivery The Problem: xygen reduction (even in 2 PEM cell) is highly inefficient wastes ~30% of energy as heat. Methanol oxidation wastes at least 25% of energy igher alcohol and hydrocarbon electrocatalysts are essentially non-existent The Goal: educe the overpotentials needed to drive each step of 2 reduction and each step of hydrocarbon oxidation. equires faster rates at lower overpotentials

10 ecent Progress Using on-oble Ion Complexes with Polymers on Carbon 2 -eduction Catalyst. Bashyam, P. Zelenay, ature 443, 63 (2006)

11 Electrocatalytic eversibility II: The oly Grail: Efficient Energy Storage Truly reversible electrocatalysts could be used to reverse the fuel cell. Generate liquid fuel when electricity is plentiful for use when it is not. n C 2 n 2 electricity (C 2 ) n 3/2 n 2

12 Electrocatalytic eversibility III: Energy- and Atom-Economical Selective Chemical Transformations Use every atom in reactants to make valuable products with minimal wasted energy. hydrogen catalyst Power e - Supply e - oxidation catalyst lefin Water ydrogen ( 2 g) graphitic carbon electrode 2 graphitic carbon electrode polymer electrolyte membrane (PEM) Glycol

13 Molecular Electrocatalysts on Carbon Electrodes Current PEM-cell electrocatalysts are noble metal nanoparticles limited ability to tailor reactivity Use coordination complexes of one or a few metal atoms strongly adsorbed to graphitic carbon For instance, copper complexes of phenanthroline adsorb strongly in water to graphite electrodes and electrocatalyze 2 reduction. Cu Lei & Anson, Inorg. Chem. 32, (1995)

14 M X X Some ther Possible Aromatic Complexes f Late Transition Metals M X X Pd Br Br u Cl M = Cu, Pd, Pt, u, etc. M Cl Cl u M L Cl Cl Cl M = Pt, Pd L = terpy, cymene I II III IV V All adsorb avidly to graphitic surfaces from water. Larger aromatic units can be appended to increase avidity ' Cl X Pd Br M = C 3, Covalent M bonding between aromatic Cl and graphite can be X Br Cl F 2 C u F 2 C CF 2 ' L CF 2 used if needed (reduction of aryl diazonium ions etc.)

15 The Ubiquitous Problem of 2 eduction: The Prospect of Cu-Based Catalysts Trinuclear Cu(I) site in laccase enzymes can reduce 2 at very positive potentials Laccase overpotential is only -70 mv Pt overpotential is -370 mv - Soukharev, Mano & eller, J. Am. Chem. Soc. 126, 8368 (2004).

16 Initial Study: ne Copper Ion Proposed Mechanism: Transfer an electron from Cu(I) to the 2 and then bind the resulting superoxide anion ( 2 ) to the Cu(II) to start reduction 2, 2 2 X Cu 4e Graphite Electrode

17 i / µa i / µα purged Me Cu 2 reduction as a function of substitution on phen ligand of Cu(I) Cu Me Air-Saturated Thermodynamic xygen eduction Potential ate of 2 eduction / s phen 3-C 2 Et-phen 3,8-(C 2 Et) 2 -phen 2,9-Me 2 -phen 2-Me-phen 2,9-Et 2 -phen S olid S tate P t C atalyst verpotential / mv Single Cu(I) site fails to bind 2 at less than 550mV of overpotential

18 Conclusion from Cu(phen) Complexes on Graphite Electrodes The overpotential of 2 reduction can be decreased by developing catalysts with electron-withdrawing substituents and with steric bulk near the Cu-center. As the overpotential of 2 reduction is decreased by these ligand effects, there is a corresponding decreases in the rate of 2 reduction. Acetic acid facilitates the 2 reduction by phen-based catalysts. When no acetic acid is present, larger overpotentials are required for 2 reduction. Future work will involve the development of multi-copper complexes that operate closer to the thermodynamic 2 -reduction potential with similar 2 -reduction rates.

19 ew Pd(II) xidation Catalyst 2 * crystal structure Pd Pd 1 2 CF 3 S 3 -

20 Dimer dissociates readily to active catalyst 2 * Pd Pd 2 CF 3 S 3 - Pd S F F F Fast initial rates of oxidation of alcohols (>0.05 s -1 ) are observed at room temperature using air as the terminal oxidant. After ~15 minutes, the rate of oxidation of 2-heptanol decreases linearly with percent conversion. Inactivation of the catalyst occurs before the reaction is complete. ne of the methyl groups is oxidized to a carboxylic acid.

21 Features of active Pd(II) oxidation catalysts Steric hindrance of dimerization At least one weak equitorial ligand Weak base to accept protons candidate being synthesized F 3 C Pd CF 3 S F F F

22 Methanol oxidation The original catalyst is a competent catalyst for the oxidation of methanol and formaldehyde in the presence of benzoquinone in acetonitrile, implicating its potential as an anode catalyst in a methanol fuel cell. Future work Immobilize catalysts on an electrode surface for use as an oxidation electrocatalyst. Synthesize fluorinated complex and test it as a catalyst for alcohol oxidation.

23 Methods to covalently site-isolate molecular catalysts on graphitic carbon nitro groups known to avoid each other graphite 3 (C3C)2 6e -, 6 electro catalyst electro catalyst electrocatalyst C= C= e -, 6 C= Cl -Cl

24 Prospects for Improved Electrocatalysts eed multiple metal ions in 2 -reduction catalysts eed appropriate proton donors and acceptors to promote rates of individual steps in catalytic cycle this sets the optimal p for each catalyst. Pd(II) with open equitorial sites attacks C- bonds and thus requires careful ligand design. eed goods ways to covalently site isolate molecular catalysts on graphitic carbon surfaces.

25 Initial work: Cu(Me 2 phen) on Graphite Electrodes otating disk voltammetry deconvolves mass transport 4 e - reduction of 2 to 2 limited by 2 binding rate (~1 s -1 ) - epeats and confirms literature results

26 Cu(II) eduction Potentials vs. 2 / 2 2 eduction Peak Potential vs. 2 / 2 2 eduction ate 2 phen Cl-phen Cl mv -925 mv -915 mv -885 mv -870 mv -935 mv -925 mv -930 mv -895 mv -905 mv 14.6 s s -1 Cl Br 2 -phen Cl Cl Br Et 2 C Br -850 mv -850 mv -805 mv -790 mv -790 mv -865 mv -865 mv -825 mv -810 mv -825 mv 5.3 s -1 Br Et 2 C C 2 Et Cl Me-phen Me Me 2 -phen Me -725 mv -640 mv -630 mv -600 mv -580 mv -745 mv /A -650 mv /A -880 mv 10.6 s s -1 Me n-bu n-bu Me 2 2 -phen Me 2 Me Me -545 mv -520 mv -505 mv -475 mv -865 mv /A /A /A

27 The First Challenge of ydrocarbon xidation: C- Activation Partial oxidation of methane and activation of other C- bonds has been explored extensively Usually, complete oxidation is to be avoided. owever, we WAT complete oxidation! Electrophilic, late transition-metal complexes are well known to insert into C- bonds Perriana, errmann and others have shown insertion of acid-stable aromatic complexes of Pt(II) and Pd(II) into C 4

28 C-C Bond Activation: Interesting Precedents PdCl 2, CuCl 2 3 C C, 2, CF 3 C 2 CF 3 3 C CF 3 3 C Figure 2. Sen s oxidative cleavage of butane. PdX 2 C, 2 PdX 2 C, 2, 2 PdCl 2, CuCl 2 C 3 Cl C, Me C 3 C 3 Figure 3. enry s carbonylation of ketones followed by retro-claisen condensation

29 Possible Strategies For C-C Bond Cleavage PdX 2-2, 2 e - PdX 2 2 PdX 2-2, 2 e -, 2 Figure 4. Proposed oxidative cleavage of long-chain carboxylic acids. u L M X u M L X - X L u M u L M M = Pd, Pd X = Cl, 2 2 u M L Pd(II) 2 2 Figure 7. Bimetallic strategy for activating C-C bonds of ketones.

30 The C xidation Challenge C poisoning of metallic Pt catalysts for 2 oxidation is well known C is expected to be a common intermediate in complete oxidation of all hydrocarbons We will explore mechanism of nucleophilic attack of 2 on Pd(II)-C and Pt(II)-C complexes u Pt 2 2 C u Pt C 2 2 u Pt 2 2 C 2 u Pt 2

31 Planned xidation Studies Investigate monometallic Pd(phen) complexes for their ability: 1. To catalyze an electrochemical version of water-gas shift reaction: C 2 -> C e - 2. To electrocatalyze simple C- activation such as alcohol oxidation.

32 Electrocatalysis of Alcohol xidation 2 2 Pd black decomposition 2 2 Pd 0 2 e- to graphite anode 0.5 Pd Pd 2 Pd Pd Pd

33 Crystal Structure of ew Alcohol xidation Catalyst 2 = * 2 CF 3 S 3 - Pd Pd

34 A Bimetallic C-C Bond Mechanism Activation: for C-C Activation Example In the mechanism below, the alcohol will be oxidized to a ketone, resulting in an activated C-C bond that can insert into the proximal metal center. This will induce oxidative cleavage of the ketone and form a carboxylic acid as one of the products. u L M X u M L X - X u L M M = Pd, Pd X = Cl, L u M 2 2 L u M

35 ich pportunity to Exploit Synergies Among Disciplines eed to understand both engineering (PEM cells) and chemistry (homogeneous catalysis) to open up new opportunities: 1. ydrocarbon fuel cell 2. Better 2 -electrodes and C-resistant 2 -electrodes for hydrogen fuel cells. 3. Electrocatalysis for energy efficient chemical processes of commodity petrochemicals particularly partial oxidations 4. Eventual electrosynthesis of hydrocarbon fuels by reverse reactions

36 eversible ydrocarbon Fuel Cell Approximate thermochemistry at 25 C and 1atm: 3 ( C2 ) n ( l) n 2( g) C ( ) (g) 2 n 2 g n 2 q rev rxn = T S rxn = n(17 kj/mol) G rxn = -n(631 kj/mol) Approximate electrochemistry at 25 C, 1atm and p 0: ( C 2) n( l) 2n 2(g) n C2( g) 6n 6n e E 0 = 0.09 V vs E 2( g) 6n 6n e 3 2(g) 3 n n 2 E 0 = 1.18 V vs E eversible cell potential = 1.09 V