clinical statistics; S. Toussi for medical monitoring; K. Bartsch for
clinical study leadership; S. Pawlak for clinical trial leadership;
C. Fredette for clinical trial project management; D. Ding for
regulatory documentation support; and M. Dolsten for scientific
discussion and advice. This research used resources of the
Advanced Photon Source, a US Department of Energy (DOE) Office
of Science User Facility operated for the DOE Office of Science
by Argonne National Laboratory under contract no. DE-AC02-
06CH11357. Extraordinary facility operations were supported in
part by the DOE Office of Science through the National Virtual
Biotechnology Laboratory, a consortium of DOE national
laboratories focused on the response to COVID-19, with funding
provided by the Coronavirus CARES Act.Funding:This study
was sponsored by Pfizer, Inc.Author contributions:
Conceptualization: C.M.N.A., A.S.A., D.R.O., M.P.; Formal analysis:
L.D., B.B., S.E.G., Q.Y., D.K.R., J.J.N., A.D., R.S.O., R.S.P.S.;
Investigation: M.A., S.B., J.C.L., J.L., K.O., L.W., R.F., K.S.G., W.L.,
R.S.O., D.K.R., S.A.G., L.A., S.N., H.E.; Methodology: J.B.T., Q.Y.,
M.F.S., P.R.V., M.R.R., M.P., D.R.O., N.C.P., A.C., E.P.K., K.J.C., R.S.
P.S.; Project administration: D.R.O., M.R.R., J.B.T., A.S.K., C.M.S.,
L.U., J.G.S., B.L.H., Y.Z., S.W.M., R.D.C., N.C.P.; Resources: A.D.,
J.J.N., K.J.C.; Supervision: M.R.R., C.E.S., A.E.S., B.L.H., P.R.V.,
R.D.C.; Visualization: S.W.M., D.K.R., S.N., B.B., L.A., H.E.; Writing–
original draft: D.R.O., A.S.K., M.F.S., C.M.S., R.D.C., M.R.R., J.G.S.,
C.M.S., Y.Z.; Writing–review and editing: D.R.O., A.S.K., M.F.S.,
A.S.A., C.M.N.A.Competing interests:D.R.O., C.M.N.A., A.S.A.,
L.A., M.A., S.B., B.B., R.D.C., A.C., K.J.C., A.D., L.D., H.E., R.F.,
K.S.G., S.E.G., E.P.K., A.S.K., J.C.L., J.L., W.L., S.N., R.S.O., K.O.,
N.C.P., D.K.R., M.R.R., M.F.S., J.G.S., R.P.S., C.M.S., A.S., J.B.T.,
L.U., P.R.V., L.W., Q.Y., and Y.Z. are employees of Pfizer and some
of the authors are shareholders in Pfizer Inc. S.W.M., J.J.N., and
M.P. were employees of Pfizer Inc. during part of this study.Data
and materials availability:All nonclinical data are available in
the main text or the supplementary materials. X-ray coordinates
and structure factors are deposited at the RCSB Protein Data Bank
under accession codes 7RFR, 7RFU, 7RFS, and 7RFW.
Individualized Fig. 5 data and the statistical analysis plan are
available on http://www.Figshare.com with the following DOI: 10.25454/
pfizer.figshare.16699669. Pfizer compound use requests are
processed through the Pure Compound Grants program (see
https://www.cybergrants.com/pfizer/Research). Pfizer Inc. has
applied for patent applications covering PF-07321332 as well as
related compounds. A phase 1 clinical trial has been registered
at http://www.ClinicalTrials.gov under identifier NCT04756531. This workis licensed under a Creative Commons Attribution 4.0 International
(CC BY 4.0) license, which permits unrestricted use, distribution,
and reproduction in any medium, provided the original work is
properly cited. To view a copy of this license, visit https://
creativecommons.org/licenses/by/4.0/. This license does not
apply to figures/photos/artwork or other content included in the
article that is credited to a third party; obtain authorization
from the rights holder before using such material.SUPPLEMENTARY MATERIALS
science.org/doi/10.1126/science.abl4784
Materials and Methods
Figs. S1 and S2
Tables S1 to S13
References ( 37 Ð 70 )
MDAR Reproducibility Checklist15 July 2021; accepted 28 October 2021
Published online 2 November 2021
10.1126/science.abl4784ELECTROCHEMISTRY
Enhancement of lithium-mediated ammonia synthesis
by addition of oxygen
Katja Li^1 †, Suzanne Z. Andersen^1 †, Michael J. Statt^2 †, Mattia Saccoccio^1 , Vanessa J. Bukas^1 ‡,
Kevin Krempl^1 , Rokas Sažinas^1 , Jakob B. Pedersen^1 , Vahid Shadravan^1 , Yuanyuan Zhou^1 ,
Debasish Chakraborty^1 , Jakob Kibsgaard^1 , Peter C. K. Vesborg^1 , Jens K. Nørskov^1 , Ib Chorkendorff^1
Owing to the worrying increase in carbon dioxide concentrations in the atmosphere, there is a need
to electrify fossil-fuelÐpowered chemical processes such as the Haber-Bosch ammonia synthesis.
Lithium-mediated electrochemical nitrogen reduction has shown preliminary promise but still lacks
sufficient faradaic efficiency and ammonia formation rate to be industrially relevant. Here, we show that
oxygen, previously believed to hinder the reaction, actually greatly improves the faradaic efficiency
and stability of the lithium-mediated nitrogen reduction when added to the reaction atmosphere in small
amounts. With this counterintuitive discovery, we reach record high faradaic efficiencies of up to 78.0 ±
1.3% at 0.6 to 0.8 mole % oxygen in 20 bar of nitrogen. Experimental x-ray analysis and theoretical
microkinetic modeling shed light on the underlying mechanism.
A
mmonia (NH 3 ) is one of the most abun-
dantly manufactured chemicals world-
wide, with a yearly production of over
182 million tonnes ( 1 ). Its main use is
as a synthetic fertilizer (~80%) and as
the source of all activated nitrogen in the
chemical industry, but it has recently also
been considered as a carbon-free energy car-
rier ( 2 – 4 ). Currently, ammonia is produced
from nitrogen and hydrogen through the ther-
mally catalyzed Haber-Bosch process, which
operates under harsh conditions (350° to
450°C, 100 to 200 bar), requiring large cen-
tralized plants and high capital investment
( 5 , 6 ). To satisfy the commercial demands,
about ~1% of global energy consumption
is used in the process ( 7 ). Additionally, the
Haber-Bosch process is responsible for about
1.4% of the annual CO 2 emissions, as the sup-
plied H 2 originates from steam reforming of
methane ( 8 – 10 ). An alternative, environmen-
tally sustainable way to produce ammonia is
through an electrochemical pathway, with the
electrical energy provided from renewable
sources such as wind or solar power. Recent-
ly, efforts toward electrochemical synthesis of
ammonia have increased substantially ( 11 – 14 );
however, the field has been hampered by var-
ious issues. One major concern in the literature
is related to contamination of the input gases,
chemicals, and catalysts by ammonia and other
labile nitrogen compounds ( 15 , 16 ), which may
result in an erroneously high reported faradaic
efficiency (FE). Several protocols ( 16 – 18 ) have
been published on proper contaminant iden-
tification and rigorous experimentation, and
some of the erroneous reports are being cor-
rected or withdrawn as scientists retest andreevaluate methods and systems ( 19 , 20 ). In
a recent paper, Choiet al. investigated over
130 publications on electrochemical ammonia
synthesis, concluding it highly likely that none
of the aqueous systems produce ammonia
and that the most reliable method presently
is lithium-mediated nitrogen reduction (LiNR)
in nonaqueous electrolytes ( 16 ). A similar
lithium-mediated method was first published
by Fichteret al.in1930( 21 )andlaterin-
vestigated with a near-aprotic solvent by
Tsunetoet al. in the 1990s ( 22 , 23 ). Currently,
the Tsuneto-based system has been revisited by
several groups ( 11 , 12 , 17 , 18 , 24 – 28 ); however,
the exact mechanism is still not fully under-
stood. It is generally accepted that the LiNR
takes place in three steps ( 25 ), the first one
being the electrochemical reduction of Li+
ions in the electrolyte to metallic Li, which is
a very reactive material. This freshly plated Li
is believed to then dissociate N 2 ,andtheN
at the surface is finally reduced in a series of
electron and proton transfers to form NH 3 by
using a suitable proton source, like ethanol
(EtOH). An important component of the LiNR
system is the solid electrolyte interface (SEI)
that forms from decomposition products of an
organic electrolyte during Li deposition on the
cathode. The SEI provides a porous passiva-
tion layer over the electrode surface that is
electronically insulating but ionically conduct-
ing. Its exact composition and mechanistic
role in the LiNR process are still unclear and
difficult to determine, because it is strongly
dependent on experimental conditions and
sensitive to air exposure. Nevertheless, the
process reliably forms ammonia from N 2
and a proton source at ambient conditions,
typically achieving a FE of around 5 to 20%
( 17 , 25 , 29 ), with a recent breakthrough in
FE of up to 69% by Suryantoet al.( 11 ) under
20-bar N 2. Lazouskiet al. implemented this
process using a gas diffusion electrode (GDE)
setup with reported efficiencies of 30%, but theSCIENCEscience.org 24 DECEMBER 2021¥VOL 374 ISSUE 6575 1593
(^1) Department of Physics, Technical University of Denmark,
Kongens Lyngby, Denmark.^2 SUNCAT Center for Interface
Science and Catalysis, Department of Chemical Engineering,
Stanford University, Stanford, CA, USA.
*Corresponding author. Email: [email protected] (J.K.N.); ibchork@
fysik.dtu.dk (I.C.)
†These authors contributed equally to this work.
‡Present address: Fritz-Haber-Institut der Max-Planck-Gesellschaft,
14195 Berlin, Germany.
RESEARCH | RESEARCH ARTICLES