1
DOI: 10.1002/ ((please add manuscript number))
Article type: Full Paper
Potassium Prussian Blue Nanoparticles: A Low-cost Cathode Material for Potassium-
ion Batteries
Chenglin Zhang, Yang Xu, Min Zhou, Liying Liang, Huishuang Dong, Minghong Wu, Yi Yang,
Yong Lei*
C. L. Zhang, H. S. Dong, Dr. M. H. Wu, Y. Yang, Prof. Y. Lei
Institute of Nanochemistry and Nanobiology, School of Environmental and Chemical
Engineering, Shanghai University, Shanghai 200444, P. R. China
E-mail: yong.lei@tu-ilmenau.de
Dr. Y. Xu, Dr. M. Zhou, L. Y. Liang, Prof. Y. Lei
Institute für Physik & IMN MacroNano (ZIK), Technische Universität Ilmenau, Ilmenau
98693, Germany
Keywords: potassium-ion batteries, Prussian blue, nanoparticles, cathode, full cell
Potassium-ion batteries (KIBs) in organic electrolytes hold great promise as an
electrochemical energy storage technology owing to the abundance of potassium, close redox
potential to lithium, and similar electrochemistry with lithium system. Although carbon
materials have been studied as KIB anodes, investigations on KIB cathodes have been
scarcely reported. We for the first time report a comprehensive study on potassium Prussian
blue K
0.220
Fe[Fe(CN)
6
]
0.805
·4.01H
2
O nanoparticles as a potential cathode material. The
cathode exhibits a high discharge voltage of 3.1~3.4 V, high reversible capacity of 73.2 mAh
g
-1
, and great cyclability at both low and high rates with a very small capacity decay rate of
~0.09% per cycle. Electrochemical reaction mechanism analysis identifies the carbon-
coordinated Fe
III
/Fe
II
couple as redox-active site and proves structural stability of the cathode
during charge/discharge. Furthermore, for the first time, we present a KIB full-cell by
coupling the nanoparticles with commercial carbon materials. The full-cell delivers a capacity
of 68.5 mAh g
-1
at 100 mA g
-1
and retains 93.4% of the capacity after 50 cycles. Considering
the low cost and material sustainability as well as the great electrochemical performances, this
work may pave the way towards more studies on KIB cathodes and trigger future attention on
rechargeable KIBs.
2
1. Introduction
The rapid increase in powering portable electronic devices and vehicles has caused a
mass production of Li-ion batteries (LIBs).
[1]
The demand of LIBs is still growing
because power storage units for large-scale stationary applications need to supplement
irregular power generation and consumption patterns so that intermittent renewable
energies can be efficiently stored and utilized.
[2]
However, the rising costs and
availability of global lithium resources have raised concerns about the heavy reliance
on LIBs because most easily accessible lithium reserves are in either remote or
politically sensitive areas.
[3]
The fact that stationary applications are indispensable for
the deployment of renewable energies and low cost upon scaling up is one of the basic
prerequisites calls for alternative Earth-abundant metal-ion batteries with similar
electrochemical principles. The abundance of sodium in Earth’s crust is three orders
higher than that of lithium,
[4]
making Na-ion batteries (NIBs) a potential candidate for
stationary applications. The redox potential of Na/Na
+
is -2.71 V (vs. SHE), which is
0.3 V more positive than that of Li, so there is an energy penalty to pay. NIBs have
received considerable attention recently, demonstrating encouraging capacity and cycle
life as well as rate capability.
[5]
In fact, potassium has a comparably high natural
abundance as sodium
[4c]
and the redox potential of K/K
+
(-2.92 V vs. SHE) is even
lower than that of Na/Na
+
, indicating a comparably low cost but even higher energy
density of K-ion batteries (KIBs). Unfortunately, very small progress has been made
on the side of KIBs so far.
KIBs came into focus very recently owing to the successful implantation of
carbonaceous materials (graphitic and non-graphitic
[6]
) as KIB anodes in nonaqueous
electrolytes. Electrochemical insertion of K-ions into graphitic carbons was reported as
a reversible three-step phase transformation process.
[6a]
Both graphite
[6b]
and reduced
graphene oxide
[6a]
exhibited significantly larger capacity (typically over 200 mAh g
-1
)
3
in KIBs than in NIBs. However, they suffer fast capacity fading and moderate
capability. Non-graphitic carbons, such as hard carbon microspheres,
[6c]
were found to
address this issue and exhibited improved cycle life and rate capability. Additionally,
metallic Sb has been used as KIB anode as well.
[7]
Despite the progress made on the
anode side, study on the cathode side hasn’t followed to the same extent, presumably
due to the larger ionic radii of K-ion (Li
+
< Na
+
< K
+
, 0.76 < 1.02 < 1.38 Å) that causes
significant restrictions for use with intercalation type electrodes. Investigations of
cathode materials of KIBs in nonaqueous electrolytes have been scarcely reported.
[8]
Lack of such investigations thus hinders the advancement of KIB full-cells.
Prussian blue (PB) and its analogues (PBAs) have been explored for many different
applications for decades because of their ease of synthesis and intriguing electrochemical and
magnetic properties.
[9]
The general chemical formula of PBAs is A
x
M[M’(CN)
6
]
1-y
·□
y
·nH
2
O
(0 < x < 2, y < 1), where A represents mobile cations, M represents nitrogen-coordinated
transition metal ions, M’ represents carbon-coordinated transition metal ions (the formula
represents PB when M = M’ = Fe), and □ represents [M’(CN)
6
] vacancy. PBAs typically
possess a face-centered cubic (fcc) crystal structure that has a three dimensional (3D) network
in which transition metal ions are linked together through cyanide (CN) ligands (Figure 1).
Each unit cell contains eight subunit cells (indicated by the red cell edges in Figure 1a) and
therefore contains eight interstitial sites that can accommodate not only neutral molecules but
ions charge-balanced by the transition ions.
[10]
Interstitial water is commonly found within the
3D network of the PBAs and have two structurally distinguishable kinds, namely zeolitic
water (1) and coordinating water (2). The former occupies the octahedral center of the
subunits and the latter chemically coordinates with the M ions because of the [M’(CN)
6
]
vacancy locating at the center of the unit cell (Figure 1b).
[11]
The 3D network, as the most
distinguish structural feature of the PBAs, provides open channels along <100> direction and
interstitial sites with the diameters of 3.2 Å and 4.6 Å, respectively, in the case of PB.
[12]
Such
4
open scaffolding leaves large cavities within, and tunnels running in three directions
throughout the roomy lattice that enables rapid solid-state diffusion of a wide variety of
intercalated ions, upon which the electrochemical properties of the PBAs can be attributed to
the redox behavior of the transition metal ions. In fact, PBAs were investigated as hosts for
alkali ions in both nonaqueous
[11a]
and aqueous electrolytes
[13]
years ago and recently regained
attention from the scientific community owing to their success in NIBs. PBAs containing
different transition metal ions (Fe,
[14]
Mn,
[15]
Co,
[16]
and Ni
[16]
) have been demonstrated as
NIB cathode materials in nonaqueous electrolytes. In aqueous electrolytes, it has been
reported that PBAs exhibited favorable electrochemical activity upon reversible insertion and
extraction of Na
+
and K
+
.
[9b,17]
To the best of our knowledge, very little concern has been
shown for PB and PBAs as KIB electrode materials in nonaqueous electrolytes. An early
work primarily showed the possibility of PB film as a KIB cathode with a mass loading only
around 80 μg,
[18]
but it received little attention presumably because of the extremely low
testing rate (~8.7 mA g
-1
), an insufficiency to demonstrate the rate performance and, more
importantly, a lack of illustrating its application in an actual KIB full-cell.
In this work, we report a comprehensive study on potassium Prussian blue nanoparticles
(KPBNPs) as a low-cost KIB cathode in nonaqueous electrolytes. The KPBNPs cathode
exhibits a high discharge voltage of 3.1~3.4 V and great cycle life at both low (73.2 mAh g
-1
at 50 mA g
-1
) and high rates (36.0 mAh g
-1
at 400 mA g
-1
), possessing a very small capacity
decay rate of ~0.09% per cycle. The electrochemical reaction mechanism study reveals that
the redox-active site of the KPBNPs is the carbon-coordinated Fe
III
/Fe
II
couple. Furthermore,
for the first time, we present an operational KIB full-cell by coupling the KPBNPs with the
commercially available carbon material Super P. The full-cell delivers a capacity of 68.5 mAh
g
-1
at 100 mA g
-1
and retains 93.4% of the capacity after 50 cycles. Considering the low cost
and material sustainability as well as the great electrochemical performances of the KPBNPs,
our work highlights the promise of PB and PBAs in the field of KIBs.
5
2. Results and Discussion
2.1. Morphology and Structural Analysis
KPBNPs were synthesized by a facile precipitation method in an aqueous solution containing
K
4
Fe(CN)
6
and FeCl
3
, in which complicated operation or tedious process is not required. The
experimental details can be found in the Supporting Information. To obtain the chemical
composition, the inductively couple plasma atomic emission spectroscopy (ICP-AES) method
and elemental analysis were used to determine the mass ratios of the heavy elements, i.e. K
and Fe, and the light elements, i.e. C, N and H, respectively (Table S1). Assuming a Fe(CN)
6
unit, the chemical composition of the as-prepared KPBNPs is determined to be
K
0.220
Fe[Fe(CN)
6
]
0.805
·□
0.195
·4.01H
2
O, which suggests the presence of a considerable amount
of [Fe(CN)
6
]
4-
vacancy. Thermogravimetric (TG) analysis shows two weight-loss events
(Figure S1): the first step occurring below 200°C corresponds to the loss of the physically
absorbed and zeolitic water, and the second step occurring above 200°C can be assigned to
the elimination of the coordinated water and the decomposition of the PB framework.
[11b,19]
The weight loss over the first step corresponds to 3.79 water molecules per formula and it is
difficult to determine the amount of the coordinated water from the second step because of the
overlap between the two occurring processes. However, by abstracting the amount of the
water determined in the first step from the total amount of the water determined in the
chemical composition, it is reasonable to estimate the amount of the coordinated water to be
0.22, which is very close to the value of the [Fe(CN)
6
]
4-
vacancy per formula in the chemical
composition within the experimental error.
Figure 2a shows the X-ray diffraction (XRD) pattern of the as-prepared KPBNPs and the
standard pattern of Fe
4
[Fe(CN)
6
]
3
(JCPDS No. 52-1907). Although there is a lack of accurate
information on K
0.220
Fe[Fe(CN)
6
]
0.805
·□
0.195
·4.01H
2
O compound in the existing
crystallographic database, it can be seen that the two patterns are almost identical to each
other. By the analogy between the obtained compound and the similar PB
[14b]
and MFe-PB (M
