HardwareX 9 (2021) e00171
Contents lists available at ScienceDirect
HardwareX
journal homepage: www.elsevier.com/locate/ohx
Hardware Article
Open source all-iron battery 2.0
Dipak Koirala, Nicholas Yensen, Peter B. Allen ⇑
University of Idaho, United States
a r t i c l e i n f o a b s t r a c t
Article history: In this work we present significant improvements to the open-source all-iron battery. We
Received 31 May 2020 show higher power density and simpler fabrication. We also show a more reproducible
Received in revised form 28 December 2020 procedure for preparing the electrolytes. The results are a highly rechargeable electro-
Accepted 28 December 2020
chemical cell based on iron, chloride, sulfate, and potassium ions in water at near-
neutral pH. The cell is stable for thousands of cycles. It displays modest energy density con-
sistent with the previous all-iron battery. The current is improved by a factor of 10 to a
Keywords:
practical level of 500 mA/L and is able to deliver a maximal power of 250 mW/L. While this
Battery
Electrochemical cell
is modest performance compared to commercial rechargeable batteries, its low cost, sim-
Rechargeable ple synthesis, and safe manufacturing may make it suitable for storing renewable energy.
Power Ó 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY
Energy license (http://creativecommons.org/licenses/by/4.0/).
Renewable
Specifications table
Hardware name Open source Iron Battery 2.0
Subject area Chemistry and Biochemistry
Hardware type Electrical engineering and computer science
Open Source License CC BY-NC-ND 4.0
Cost of Hardware $4.58 - $13.74 for cells of 8 ml each; Cost of set-up $ 400 (to build 50 cells of 8 ml each)
Source File Repository https://doi.org/10.17605/OSF.IO/YV2E6
1. Hardware in context
Inexpensive, safe energy storage has many applications. Renewable energy can only displace a percentage of fossil fuel
energy unless it can be efficiently and cost-effectively stored [1]. Lithium-ion batteries have emerged as the dominant energy
storage system for mobile applications, but they have safety [2] and cost issues [3]. For stationary applications, it may be
advantageous to move to a cheaper, safer, but lower energy density chemistry. We demonstrated a small-scale all-iron bat-
tery [4]. This cell was highly rechargeable with modest but useful energy density suitable only for low-power applications.
We here report an improved version of this chemistry with similar energy density and much higher power density as well as
a more convenient form factor.
⇑ Corresponding author.
E-mail address: [email protected] (P.B. Allen).
https://doi.org/10.1016/j.ohx.2020.e00171
2468-0672/Ó 2021 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
D. Koirala, N. Yensen and P.B. Allen HardwareX 9 (2021) e00171
Several all-iron batteries have been demonstrated in the literature and at large scale for commercial applications. For a
recent summary, see Anarghya et al. [5]. Such batteries are often implemented as flow batteries [6,7]. Flow batteries have the
advantage of decoupling the energy capacity (determined by the size of the electrolyte tank) from the power capacity (de-
termined by the size of the flow cell). This comes at the cost of relatively complex plumbing and pumps.
The all-iron battery presented here is a conventional battery and not a flow battery. Although the chemical reactions that
move and store electrons are the same (i.e., the oxidation of Fe and the reduction of Fe3+), the physical design is much sim-
pler. Rather than using a high-performance flow cell to achieve practical power levels, our approach is to modify the anode
and cathode materials to achieve useful energy and power performance.
2. Hardware description
We describe a design for an energy storage battery with an iron-based anode and cathode. The overall strategy is shown
in Fig. 1. Iron metal is oxidized to ferrous iron at the anode while ferric iron is reduced to ferrous iron at the cathode allowing
electrons to flow. This system provides high rechargeability and customizability, but the power and energy density are low
compared to commercial batteries. We also describe the simple synthesis of a suitable separator membrane made from cel-
lulose acetate impregnated paper. This separator keeps the anode and cathode chemistry separate but allows ion transfer.
The area of the polymer sheet determines the maximum current. For more power, the battery should be thin with a large
surface area. If less power and more energy capacity is needed, the battery can be thicker with a smaller separator membrane
area (i.e., a thicker acrylic plastic or multiple layers can be used).
Our primary improvement to the original open-source all-iron battery is to increase the current density. The original for-
mulation [4] was a mix of iron chloride and potassium sulfate adjusted to pH 7.5 with sodium hydroxide. Raising the pH
causes a solid to precipitate. The conductivity of the precipitated particles is low and the concentration of iron ions in solu-
tion is also low. This limited the battery’s maximal electrical current. We determined that the maximal discharge and charge
current could be significantly increased by adding conductive carbon. By decreasing the distance between the carbon elec-
tron conductor (originally, carbon felt) and the iron precipitate, we were able to increase power density. Adding ketjen black
Fig. 1. Overall battery design and function. (A) Images show the three active forms of iron. (B) A schematic show how the overall battery construction and
current flow.
2
D. Koirala, N. Yensen and P.B. Allen HardwareX 9 (2021) e00171
Fig. 2. Ketjen black conductive particles improve battery performance. (A) Schematic shows how electrons diffuse through electrolyte. (B) Schematic shows
how hypothetical effect of conductive carbon to shorten diffusion distance. (C) Graph of Sustained current vs % (w/w) of conductive carbon.
[Conductive Carbon] to the anode and cathode pastes allows electrons to migrate to within a much shorter distance of the
active material through a conductive matrix (see Fig. 2) We added a range of concentrations of Ketjen black [conductive car-
bon] to the precipitate and determined that 4% carbon by mass was sufficient to reach near-optimal performance. More than
4% carbon means that there is less room for iron and so higher concentrations are not recommended.
As a secondary improvement we sought to simplify the original iron battery chemistry. In the original chemistry, the iron
salt solutions contained dissolved iron, sodium, potassium, chloride and sulfate ions. This is shown at left, using the previous
best membrane. The result was a stable chemistry with a large voltage drop (Fig. 3, far left). We replaced the sodium poly-
acrylate membrane with an improved membrane (cellulose acetate drop cast on printing paper). This improved the power
density resulting in less internal resistance and a smaller voltage drop during discharge. We also systematically eliminated
each component to determine which were necessary. Sulfate, chloride and potassium ions are necessary to generate a high
performance, rechargeable cell. Only the removal of the sodium ions was tolerated with high performance (see Fig. 3, No
Na+). Eliminating any of the other ions resulted in cell degradation and a changing charge–discharge profile. The newest for-
mulation includes iron chloride and potassium sulfate precipitated with potassium hydroxide. Relative to the original chem-
istry, the new formulation is slightly simpler. However, depending on availability of materials, other formulations are
possible (e.g., the use of ammonium iron sulfate) so long as the appropriate ions are dissolved in solution prior to
precipitation.
3
D. Koirala, N. Yensen and P.B. Allen HardwareX 9 (2021) e00171
Fig. 3. Optimization of cell chemistry includes cyclic potentiograms of each of four formulations with original iron chemistry (Iron Battery 1.0, at far left). In
each case one ionic component (Na+, K+, Cl, or SO2–
4 ) was eliminated. The graphs show potential as a function of time during the first 10 and last 10 of 500
total cycles. Consistency in the charge-discharge profiles indicates stability.
This hardware could be useful in any context where a safe, inexpensive battery would be an advantage. The ability to use
an open source energy storage solution could complement open source hardware projects in many fields:
Storage of renewable energy from an open source wind turbine [8]
Storage of renewable energy from an open source solar farm [9]
Open source environmental data loggers [10]
Open source weather stations [11]
3. Design files
Design Files Summary:
Design file name File type Open source license Location of the file
Battery Design CAD file in .svg format CC BY-NC-ND 4.0 doi.org/https://doi.org//10.17605/OSF.IO/YV2E6
Battery Design: The design files are included online at https://doi.org/10.17605/OSF.IO/YV2E6. The specified design
includes a cell with external dimensions of 165 mm 85 mm 5 mm and an internal active volume of 9.6 ml. The cell
is assembled from four acrylic parts and held together by a combination of solvent welding, glue and machine screws.
4. Bill of materials
The following is for 3 cells containing 8 ml each of the electrolyte pastes.
Designator Component Number (per 3 Cost per 8 ml Cost per Total Source of Material
cells of 8 ml cell - USD ($) unit - USD cost - materials type
each) ($) USD ($)
Fe Salt 1 Ferrous 3.97 g $0.25 $0.19/g $47.90 Alfa-Aesar Inorganic
Chloride $1.11 $0.84/g $21 (A16327)
(FeCl2) $0.66 $0.50/g $49.95 CPLabSafety
Amazon
(B00QG9I3SK)
Fe Salt 2 Ferric 3.24 g $0.08 $0.08/g $39.10 Alfa Aesar Inorganic
Chloride $0.04 $0.035/g $17.60 (A16231)
(FeCl3) $0.14 $0.13 $12.95 Amazon
(B00DYOA85Q)
ebay
4
D. Koirala, N. Yensen and P.B. Allen HardwareX 9 (2021) e00171
⇑ (continued)
Designator Component Number (per 3 Cost per 8 ml Cost per Total Source of Material
cells of 8 ml cell - USD ($) unit - USD cost - materials type
each) ($) USD ($)
Salt Potassium 6.97 g $0.08 $0.034/g $34.40 Alfa Aesar Inorganic
Sulfate $0.23 $0.10/g $5.09 (A13975)
(K2SO4) ebay
Base Potassium 8.42 g $0.13 $0.047/g $23.70 Alfa Aesar Inorganic
Hydroxide $0.42 $0.15 $15.31 (A18854)
(KOH) $0.28 $0.10/g $5.09 Amazon
(B07JVVTP56)
ebay
Fe Metal Steel Wool 0.44 g $0.01 $0.02/g $3.78 Amazon/ Metal
(Fe) Walmart
(B074MDTWQR)
Cellulose 0.45 g $0.07 $0.48/g $11.92 Fisher-Scientific Organic
Acetate (AC177780250)
Nafion 45 mL $0.04 $2.46/ml $61.50 Alfa Aesar Organic
(42118)
Ethylene 67.5 mL $0.01 $0.074/g $18.60 Alfa Aesar Organic
Glycol (A11591)
Printing 1 pcs $0.01 $0.013/ $6.44 Amazon Composit
Paper sheet (B0050MRBA0)
Conductive Ketjen Black 1.20 g $0.43 $1.07/gm $60 Ebay Inorganic
Carbon EC-600JD $0.52 $1.31/gm $65.50 Ebay
Housing Arcylic sheet 6 pcs (500 700 ) $4.40 $2.20/pcs $21.99 Amazon Composite
Or, 1700 cm2 $1.7 $0.003/cm2 ~$30 (B081B15HL4)
Local Hardware
Store
Glue Clear Seal ~6 ml $0.05 $0.025/ml $3.97 Amazon/ Composite
Walmart
(B001T8UDOU)
Elec 1 Graphite 450 cm2 $0.15 $0.001/ $11 Alibaba Semi-
sheet $1.12 cm2$0.008/ $23 Amazon conductor
cm2 (B07K8Y4269)
Methylene 6 ml $0.11 $0.057/ml $28.40 Alfa Aesar Organic
Chloride (L13089)
Copper tape 30 cm2 ~$0.50 – $2.63 Walmart Metal
Nuts and 30 pcs $5 $0.5/pc $15 Walmart Metal/
bolts $1 $0.1/pc Local hardware Alloy/
store Plastics
Totals (Lowest) $4.58/Cell (8 ml)
Totals (Highest) $13.74/ Cell (8 ml)
*Above mentioned prices are exclusive of local tax and shipping.
5. Build instructions
Short build instructions are as follows (detailed build instructions are included as Supplemental material). The cell hous-
ing was laser-cut from acrylic plastic (using a suitable CO2 laser cutter such as our BossLaser 80 W or acquired from a com-
mercial service like Ponoko). The design can be adjusted to a wide range of dimensions. The CAD designs for the laser cut
acrylic plastic (PMMA) are shown in Fig. 4A. Two copies of each part should be made. The internal enclosed volume deter-
mines the energy storage capacity. The surface area determines the power.
Graphite foil [Elec1] was cut to shape with tabs for making an electrical connection with alligator clips. The graphite foil
was glued to a flat acrylic sheet using clear seal [Glue]. A second acrylic sheet was cut with a central hole slightly smaller
than the graphite foil and glued to the first. The design leaves an open cavity for active material (i.e., either anode electrolyte
paste or cathode electrolyte paste) as illustrated in Fig. 4B. A separator membrane was prepared by dissolving cellulose acet-
ate in acetone, applying the solution to paper along with an optional small volume of Nafion, and allowing to evaporate (see
Fig. S.2).
5
D. Koirala, N. Yensen and P.B. Allen HardwareX 9 (2021) e00171
Fig. 4. Cell design and construction. (A) CAD design of cell housing. (B) Schematic of cell assembly. (C) Photograph of image of materials and selected steps
of cell assembly.
Electrolyte paste was prepared by dissolving iron chloride (ferric for cathode and ferrous for anode) in water at the appro-
priate proportions. To this, a measured quantity of potassium hydroxide solution was added to reach pH 7.5. To the resulting
precipitate, ball-milled ketjen black [conductive Carbon] was added to reach 4%. The anode cavity was filled with finely
divided steel wool [Fe Metal] and then packed with anode paste. The separator membrane was placed over the anode cavity.
The cathode cavity was filled with cathode paste, then the cathode assembly was pressed over the separator membrane. The
full assembly was then fastened together with machine screws at the perimeter and sealed by solvent welding the acrylic
plastic with methylene chloride. Copper tape was used to reinforce the graphite foil tabs which extend beyond the edge
of the acrylic. As assembled, the cell is in the charged state and is ready to generate electrical current.
6. Operation instructions
Once the battery is assembled it can be discharged and charged like any battery or electrochemical cell. This can generate
the necessary electric current for a given electronic device. For discharge, the battery can be used to power any direct current
electrical device that draws less than the maximum current for the cell (4.5 mA for the 8 ml cell). Multiple cells can be wired
together in series or in parallel to generate higher voltage or current. For charging, we recommend a constant, regulated
charging voltage of 1.1 V per cell in series.
As one example application, we used a single cell with a ‘‘Joule thief” circuit to boost the voltage. This circuit converts the
low voltage DC to AC and drives the AC voltage through a step-up transformer. The result is sufficient for a modest brightness
LED (see Fig. 5). As an alternative, five such batteries could be wired in series. They could act as a storage reservoir for solar
energy to operate a light at night. This battery may represent some advantages over lithium ion batteries including lower
cost and lower environmental impact of production as well as low toxicity and high recyclability.
The iron battery has significant internal resistance. Beyond its maximal current, the internal resistance results in a signif-
icantly lower operational voltage. The lower voltage reduces the available power. The maximal energy density for the 8 ml
cell was 0.25 mW per ml. A battery should be designed and constructed with this limitation in mind. Depending on power
requirements, a battery of the appropriate size will be needed.
During the preparation of reagents, there are several safety considerations. Potassium hydroxide is caustic and should not
come into contact with skin. Ferric chloride is corrosive and strongly acidic and should not be allowed to come into contact
with skin. Additionally, ferric chloride will degrade metals it contacts. Once the components are neutralized by mixing in the
proper proportions, they are comparatively inert. When the cell is no longer needed, it can be considered non-hazardous
6
D. Koirala, N. Yensen and P.B. Allen HardwareX 9 (2021) e00171
Fig. 5. Iron battery 2.0 and Joule thief power an LED. (A) Photograph shows open circuit (LED off). (B) Photograph shows closed circuit (illuminated LED).
waste (subject to local regulations). The plastic, neutral pH water, iron salts, and potassium salts are all non-hazardous waste
(comparable to normal food and household waste).
7. Validation and characterization
The 8 ml version of the iron cell was tested for its total capacity and maximal power. The cell is shown in Fig. 6A. This cell
contains 4 ml each of anode and cathode paste. It has a membrane of 48 cm2. To determine the total capacity, the cell was
discharged at a constant current of 1 mA until the potential dropped to 75 mV as shown in Fig. 6B. The Total capacity on first
discharge was 80 mAh. With 0.97 g of ferric iron acting as a cathode, we would expect a maximum of 160 mAh. Thus, we are
Fig. 6. Cell characterization. (A) Photograph of assembled cell. (B) Graph shows deep discharge before and after 1000 cycles. (C) Graph of power vs current
shows maximal power.
7
D. Koirala, N. Yensen and P.B. Allen HardwareX 9 (2021) e00171
accessing 50% of the iron in the cathode. After 1000 cycles, the capacity was re-measured and was 72 mAh, showing a loss of
only 10%.
To determine the maximal power, the 8 ml cell was discharged at a range of currents and the voltage was measured. The
power was calculated at each condition by multiplying the current and voltage. The result was a graph of power as a function
of current, as shown in Fig. 6C. The maximal power was ~ 2 mWatts or 0.25 mW/ml at 4.5 mA. This is significantly lower than
typical commercial batteries, but nearly two orders of magnitude larger than our previous best.
Capabilities and limitations:
Voltage: Although the cell shows an open cell potential of 1.1 V, the voltage may drop as low as ~ 200 mV during high
current discharge. This internal resistance reduces available power.
Volumetric Capacity: 10 Ah/L. This limits the battery to stationary applications.
Energy density: >3 Wh/L
Power density: 250 mW/L
Cycle stability: Stable for greater than 1000 cycles
Price: $60 per watt-hour for 8 ml cells (as built, including laser-cut housing)
Price: $0.36 per watt-hour for active materials at wholesale prices
Human and animal rights
This work did not use human or animal subjects.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.ohx.2020.e00171.
References
[1] C. Budischak, De A. Sewell, H. Thomson, L. Mach, D.E. Veron, W. Kempton, Cost-minimized combinations of wind power, solar power and
electrochemical storage, powering the grid up to 99.9% of the time, J. Power Sources 225 (2013) 60–74, https://doi.org/10.1016/j.
jpowsour.2012.09.054.
[2] B. Liu, Y. Jia, C. Yuan, L. Wang, X. Gao, S. Yin, J. Xu, Safety issues and mechanisms of lithium-ion battery cell upon mechanical abusive loading: A review,
Energy Storage Mater. 24 (2020) 85–112, https://doi.org/10.1016/j.ensm.2019.06.036.
[3] M. Naumann, R.C. Karl, C.N. Truong, A. Jossen, H.C. Hesse, Lithium-ion battery cost analysis in PV-household application, Energy Procedia 73 (2015) 37–
47, https://doi.org/10.1016/j.egypro.2015.07.555.
[4] N. Yensen, P.B. Allen, Open source all-iron battery for renewable energy storage, HardwareX 6 (2019) e00072, https://doi.org/10.1016/j.ohx.2019.
e00072.
[5] A. Dinesh, S. Olivera, K. Venkatesh, M.S. Santosh, M.G. Priya, Inamuddin, A.M. Asiri, H.B. Muralidhara, Iron-based flow batteries to store renewable
energies, Environ. Chem. Lett. 16 (3) (2018) 683–694, https://doi.org/10.1007/s10311-018-0709-8.
[6] K.e. Gong, F. Xu, J.B. Grunewald, X. Ma, Y. Zhao, S. Gu, Y. Yan, All-soluble all-iron aqueous redox-flow battery, ACS Energy Lett. 1 (1) (2016) 89–93,
https://doi.org/10.1021/acsenergylett.6b00049.s001.
[7] A.K. Manohar, K.M. Kim, E. Plichta, M. Hendrickson, S. Rawlings, S.R. Narayanan, A high efficiency iron-chloride redox flow battery for large-scale
energy storage, J. Electrochem. Soc. 163 (1) (2016) A5118–A5125, https://doi.org/10.1149/2.0161601jes.
[8] D. Bartmann, D. Fink, Homebrew wind power: A hands-on guide to harnessing the wind, Buckville Publ. LLC (2009).
[9] A.J. Buitenhuis, J.M. Pearce, Open-source development of solar photovoltaic technology, Energy Sustain. Develop. 16 (3) (2012) 379–388, https://doi.
org/10.1016/j.esd.2012.06.006.
[10] E.d. Baker Open source data logger for low-cost environmental monitoring BDJ 2 e1059 10.3897/BDJ.2.e1059.suppl2.
[11] G.T. Netto, J. Arigony-Neto, Open-source automatic weather station and electronic ablation station for measuring the impacts of climate change on
glaciers, HardwareX 5 (2019) e00053, https://doi.org/10.1016/j.ohx.2019.e00053.
