Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
SUPPLEMENTARY INFORMATIONDOI: 10.1038/NMAT3486
NATURE MATERIALS | www.nature.com/naturematerials 11
Supporting Information on “A Rechargeable
Room-Temperature Sodium Superoxide (NaO2)
Battery“
Pascal Hartmann1, Conrad L. Bender1, Miloš Vračar1,2, Anna Katharina Dürr3, Arnd
Garsuch3, Jürgen Janek1* and Philipp Adelhelm1*
1 Physikalisch-Chemisches Institut, Justus-Liebig-Universität Gießen, Heinrich-Buff-Ring 58,
35392 Gießen, Germany
2 current address: BELLA, Institut für Nanotechnologie, Karlsruher Institut für Technologie,
Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany
3 BASF SE, 67056 Ludwigshafen, Germany
* Corresponding Authors: Jürgen Janek, [email protected]; Philipp
Adelhelm, [email protected]
A RECHARGEABLE ROOM-TEMPERATURE SODIUM
SUPEROXIDE (NAO2) BATTERY
© 2012 Macmillan Publishers Limited. All rights reserved.
2
Table S1: Thermodynamics, capacity and energy density of possible reactions in Li/O2 and Na/O2
cells. LiCoO2, a classic cathode material used in lithium-ion batteries, is added for comparison.
Specific capacity and theoretical energy density values are given on basis of the discharge product
mass.
Discharge
product Reaction
rS /
Jmol-1
K−1
rH /
kJ mol-1
rG /
kJ mol-1
E° /
V
z
Spec.
capacity /
mAhg−1
Theor. energy
density /
Whkg−1
Na2O 2 Na + ½ O2 Na2O -130.14 -415.09 -376.30 1.95 2 867.38 1691.38
Na2O2 2 Na + O2 Na2O2 -212.94 -513.21 -449.72 2.33 2 689.00 1605.37
NaO2 Na + O2 NaO2 -140.55 -260.66 -218.76 2.27 1 488.21 1108.22
Li2O 2 Li + ½ O2 Li2O -123.20 -597.88 -561.15 2.91 2 1793.88 5220.19
Li2O2 2 Li + O2 Li2O2 -206.89 -632.50 -570.82 2.96 2 1168.33 3458.26
LiCoO2
Li0.5CoO2 + 0.5 Li
LiCoO2
~4* 0.5 136.92 547.68
Thermodynamic data (T = 298 K) taken from HSC Chemistry 7.14 database, ©
Outotec
Research Center.
Values for E° are calculated from the Gibbs energy of the cell reaction E°= −rG/(zF), i.e.
the E° refers to a metal Me/Me+ anode. z is the number of electrons transferred in the reaction.
F is the Faraday constant. Specific capacities and energy densities are given per weight of
discharge product.
* Ref: Handbook of Battery Materials, Wiley-VCH, 2
nd edition, 2011
Remark on the thermal stability of NaO2:
Sodium superoxide (NaO2) can decompose to Na2O2 and Na2O upon heat treatment.
Measurements on the thermal stability indicate a thermal stability up to 160 °C (J. A. Marriott
et al., Thermochim. Acta, 2, (1971), 135-141)
© 2012 Macmillan Publishers Limited. All rights reserved.
3
Figure S1 – Galvanostatic measurements on Na/O2 and Li/O2 cells. a) a Na/O2 cell with diglyme/NaOSO2CF3 electrolyte and low surface area GDL cathode (1 m²/g) discharged at 40 µA/cm². The yielded charge output corresponds to a discharge capacity of 700 mAh/g (11 mg cathode mass). b) Li/O2 cell with diglyme/ LiOSO2CF3 and high surface area cathode comprised of 90 wt% carbon black (Super-P Li, 60 m²/g) and 10 wt% PvDF binder. Cycled at 63 µA/cm², the discharge capacity per carbon black mass is approx. 3000 mAh/g. c) Na/O2 battery with ethylene carbonate/dimethyl carbonat/NaPF6 electrolyte and a high surface area cathode as described in b). Cycled at 40 µA/cm², the discharge capacity per carbon black mass is approx. 400 mAh/g. The notch at the beginning of the charging cycle is due to a power breakdown of the potentiostat.
© 2012 Macmillan Publishers Limited. All rights reserved.
4
Figure S2 – Cyclic voltamogram of the sodium/oxygen cell. The peaks corresponding to the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) are close to the standard potential of NaO2 formation from Na and O2 (E
0 = 2.27 V). The geometric area of the electrode was
1.13 cm2. The arrow indicates the direction of a small shift of the oxidative peak upon cycling. (TfONa
= sodium triflate, NaOSO2CF3).
The curve is comparable to CVs in the literature reporting on the formation of superoxide anions in aprotic solvents. See i) Laoire, C. O., Mukerjee, S., Abraham, K. M., Plichta, E. J. & Hendrickson M. A. Elucidating the Mechanism of Oxygen Reduction for Lithium-Air Battery Applications. J. Phys. Chem. C 113, 20127-20134, (2009). ii) Laoire, C. O., Mukerjee, S., Abraham, K. M., Plichta, E. J. & Hendrickson M. A. Influence of Nonaqueous Solvents on the Electrochemistry of Oxygen in Rechargeable Lithium-Air Battery. J. Phys. Chem. C 114, 9178-9186, (2010).
© 2012 Macmillan Publishers Limited. All rights reserved.
5
Figure S3 – a Capacity evolution, discharge and charge capacity as function of the cycle number, of two sodium/oxygen cells at different current densities. The inlet shows the charge efficiency, here
defined as Q = (Qchg/Qdis)100%, as function of the cycle number. b Discharge-charge characteristics of the cell cycled at 200 µA/cm
2.
As can be seen, the efficiency during first charge is lower compared to the subsequent cycles. Eventually, this can be related to a first activation of the electrode, i.e. reaction of the electrolyte or NaO2 with the surface of the carbon cathode.
© 2012 Macmillan Publishers Limited. All rights reserved.
6
Figure S4 – Open circuit potential (EOC) of a Na/O2-cell (i) before and (ii) during oxygen flushing of the cathode, and (iii) relaxation after discharge of the cell to 1.8 V: Right after oxygen exposure, the Eoc instantly increases to about 2.27 V - 2.30 V. However at this early stage of reaction, the system can be far away from equilibrium due to the lack of enough reaction product. After discharge, the activities of the reactants should be close to unity and the measured Eoc should be closest to the standard potential E
0 which is given in Table S1 for different reaction products. The measured
potential of 2.25 V agrees well with the standard potential for a cell forming NaO2 (E0=2.27 V).
© 2012 Macmillan Publishers Limited. All rights reserved.
7
Discussion on the maximum capacity achievable capacity
Based on the dimensions of the carbon cathode (1.13 cm2 x 0.21 mm) and the porosity of
about 80 %, complete filling of the pores (0.019 cm³) with NaO2 ( = 2.2 g/cm3) should result
in a discharge capacity of approx. 73.2 C or 20.3 mAh (0.0418 g or 0.0076 mol). This is the
maximum theoretical capacity that can be achieved with this type of carbon cathode. In our
cell, we achieved a maximum discharge of 7.5 mAh at 40 µA/cm² (Figure S1), which
corresponds to 37 % of the maximum theoretical capacity.
For the following reasons, the practicable achievable capacity is lower than the theoretical
capacity:
1) In our cell (Swagelok type), the electrodes and the separator are fixed between a stainless
steel filter disk and the cell housing by a spring (see Figure S5). This leads to a compression
of the flexible carbon fiber cloth. It can be shown that the cathode volume is reduced by 20 %
(force dependent thickness measurements revealed an even higher value), consequently
decreasing the maximum theoretical capacity to around 16 mAh, i.e. the 7.5 mAh at achieved
at 40 µA/cm² correspond to almost 50 % of the theoretical capacity.
2) A stainless steel filter disk is used to allow oxygen diffusion towards the cathode. We
found that NaO2 formation preferentially occurs directly at the holes of the filter discs,
blocking access of oxygen to the whole cathode volume. This phenomenon can be seen from
the Figure 3a and Figure S5. The left, dense part covered with discharge product had direct
access to the oxygen environment, whereas the right, less dense part was covered by the filter
disk. Consequently NaO2 formation does not occur homogeneously throughout the whole
cathode structure, limiting the practical capacity achievable.
3) The cathode is immersed in liquid electrolyte and therefore the growing discharge product
(NaO2) has to displace the electrolyte from the voids. Taking to calculated pore volume
(0.019 cm3) the cathode is filled with electrolyte to at least 90 %. Displacement of the
electrolyte might be non-ideal, i.e. the practicable accessible volume for NaO2 formation is
reduced.
4) As also mentioned in the manuscript it is a well-known fact, that the transport of oxygen
within the electrolyte filled carbon structure is a rate limiting process. The discharge product
is preferentially build at interfaces direct in contact to the gas atmosphere while a depletion of
oxygen inside the cathode structure occurs during discharge. This is well in line with the
findings of our SEM investigations as shown in Figure 3 and Figure S5.
© 2012 Macmillan Publishers Limited. All rights reserved.
8
In light of these arguments, it becomes clear that an increase in practicable capacities can be
achieved by suitable cell engineering and cathode structure design.
Figure S5 – Detail sketch of the electrode assembly and the oxygen support (left). Magnified SEM image of Figure 3a (right). The area with a high amount of discharge product was located directly under a hole of the filter disk and therefore in direct contact with the oxygen atmosphere. The part with a low concentration was covered by the filter disk.
© 2012 Macmillan Publishers Limited. All rights reserved.
9
Figure S6 – (a) XRD patterns of a discharged cathode and reference patterns of commercial chemicals and the used sample holder setup. Reference material of NaO2 was kindly provided by M. Jansen (MPI-FKF, Stuttgart, Germany). The diffraction patterns of NaO2 and Na2O appear similar, as
both compounds show a cubic crystal symmetry (Fm ̅m) and only slightly differing lattice constants of 5.55 Å and 5.51 Å, respectively. This leads to similar positions of the reflections. However, obvious
discrepancies in the relative intensities of the reflections exist (see reflections at 2 = 32.5° and 46.5°) that allow distinguishing between NaO2 and Na2O. One could argue that still small amounts of Na2O might be present, but the results from Raman spectroscopy evidence phase purity of NaO2. (b) Position of reflections of NaO2 and Na2O according to JCPDS pdf database.
© 2012 Macmillan Publishers Limited. All rights reserved.
10
Figure S7 – Sketch of the cell design. Three electrode setup, sodium metal as reference electrode, encapsulated oxygen reservoir, pressure transducer (Omega engineering) for on-line measurement of the oxygen pressure at the cathode side.
© 2012 Macmillan Publishers Limited. All rights reserved.
11
Current densities
Table S2: Different practical parameters for quoting applied currents and achieved capacities
are used in literature. We calculated the most common ones for our cell and compared them
with a report from literature on a Li/O2 cell using a high surface area carbon cathode.
Numbers in italic are calculated on the basis of values given in the cited publication.
Na/O2 (this work) Li/O2 (Shao-Horn et
al.)**
negative electrode type carbon fiber cloth carbon black
material Freudenberg GDL
(H2315)
Vulcan Carbon XC72
carbon mass
(mg) 11 0.5
electrode area
(cm2)
1.13 1.26
spec. surface area
(m2/g)
< 1 100
total surface area
(cm²) 100 500
applied currents
(mA) 0.13 - 0.57 0.05 - 1.0
geometric current density
(mA/cm²) 0.12 - 0.50 0.04 - 0.79
local current density
(µA/cm²) 1.3 - 5.6 0.1 - 2.0
gravimetric current density
(mA/g) 12 - 52 100 - 2000
max. spec. discharge capacity
(mAh/g) 300 3000
amount of discharge product*
(µmol) 119 (NaO2)* 28 (Li2O2)*
total surface area = (electrode mass) •(spec. surface area)
geometric current density = (applied current)/(electrode area)
local current density = (applied current)/(total surface area)
gravimetric current density = (applied current)/(carbon mass)
* For our cell the discharge product is NaO2, for the literature cell it is claimed to be Li2O2
respectively. Therefore it has to be mentioned that in one mol peroxide twice the amount of
charge is stored compared to the superoxide.
** Shao-Horn et al. The discharge rate capability of rechargeable Li-O2 batteries, Energy
Environ. Sci. 4, 2999-3007, (2011).
© 2012 Macmillan Publishers Limited. All rights reserved.