Electrical energy storage modules offer a solution to the critical problem of inefficient microgrid systems in remote locations, such as the U.S Army’s Forward Operating Bases (FOBs). These currently rely on diesel-powered generators for primary and backup energy supply and are often underrun (i.e., operate at 40%–50% of their peak output capabilities) to accommodate power consumption fluctuations. This operating mode significantly reduces their efficiencies and drives up diesel consumption [1
Lithium-ion batteries (LIBs) offer sufficiently high energy densities to act as energy management systems in microgrid power applications [2
]. Typical batteries implement a graphite anode and a metal oxide cathode (such as LiCoO2
), along with an electrolyte composed of lithium salts (typically [Li+
], or [Li+
]) dissolved in carbonate solvents [3
]. Single cells typically operate at 3.6 V and can store 700 to 2400 mAh per cell. The LiCoO2
cathode has a theoretical gravimetric capacity of 274 mAh/g (140 mAh/g practical), while the anode is capable of storing up to 372 mAh/g (300–320 mAh/g practical) [2
]. Nevertheless, existing LIBs feature many drawbacks that preclude their implementation in grid-level storage. High-voltage output is limited by the electrochemical stability of electrolytes and requires an inefficient series connection of multiple modules [4
]. Oxidized cathodes facilitate gas evolution during electrolyte breakdown, pose swelling/bursting risks, and increase flammability hazards [5
]. The cathode intercalation compound, Lix
<0.5), contributes an additional Co4+
oxidizer hazard. Despite recent improvements of operating efficiencies, charge storage decays with cycling and offers a limited operating life. In addition to including rare, expensive transition metals such as cobalt, expended LIBs cannot be safely disposed of or incinerated without generating toxic byproducts [7
A promising alternative LIB approach depends on a dual-intercalation mechanism that relies on both cations and anions [8
]. It allows inexpensive and environmentally safe graphite to act as both the cathode and anode. While [Li+
] anode intercalation still yields LiC6
], cathode intercalation produces (PF6
]. This process, which occurs at 4.5–5.2 V, has a theoretical intercalation capacity of 90–120 mAh/g [11
]. A mixture of monofluoroethylene carbonate (FEC) and ethylmethyl carbonate (EMC), along with a tris(hexafluoro-iso-propyl)phosphate (HFIP) sacrificial additive [11
], acts as an electrochemically stable electrolyte solvent. This approach allows the use of less volatile solvents and inexpensive, easy-to-dispose electrodes, and offers energy densities that are suitable for grid-level implementation. Furthermore, the dual-intercalation process has broad implementation across a wide range of materials, including layered carbides [13
], polymer composites [14
], graphene [15
], and aluminum [16
Despite important advantages in the dual-intercalation charge storage mechanism, existing all-graphite battery configurations do not yet yield sufficiently high cycling efficiencies and steady-state discharge capacities. Furthermore, although extensive studies of solid/electrolyte (SEI) formation and stability on graphite anodes, as well as reactions involving LiCoO2
cathodes, have been completed, SEI formation on graphite cathodes due to anion intercalation has not been examined. Since the intercalation process strongly depends on processes at the electrode–electrolyte interface [17
], heterogeneity and composition differences on electrode surfaces are expected to significantly influence capacity and cycling efficiencies. Certain functional groups can be introduced onto carbon surfaces via high-temperature treatments, air oxidation, plasma and gas treatment, and wet chemistry reactions [18
]. Subsequently, research efforts must classify surface chemical species as beneficial or detrimental to anion intercalation, determine optimal combinations of cathode and anode surface treatments, and implement them in the graphite cathode fabrication processes.
Our efforts determined the influence of hydrogen and oxygen surface functionalities on the electrochemical performance of graphite cathodes and anodes in dual-ion cycling configurations. Since surface modification efforts have different effects on the graphite basal planes and sheet edges, we relied on both spherical (core-shell MCMB (MesoCarbon MicroBeads) cathodes and CGP (ConocoPhillips Graphite Particles) anodes) and flake-like SFG (Synthetic Flakes of Graphite) materials; all particle morphologies had been previously implemented in dual-intercalation LIBs [11
]. We relied on conventional and industrially scalable furnace (flowing air at 570 °C or H2
gas at 800 °C to, respectively, oxidize or hydrogenate surfaces) methods and combined characterization of surface compositions with resulting electrochemical performance.
We decoupled the surface chemistry effects for cation and anion intercalation and matched optimal surface treatments of positive and negative electrodes in asymmetric configurations to extract the highest capacities and most stable cycling behaviors.
3. Materials and Methods
All graphite materials were used as received. The first cathode material was a core-shell graphitized particle (hard carbon surface coating surrounded well-ordered graphite) MCMB 10-28 (Osaka Gas, Japan). The second cathode material was a synthetic graphite flake material SFG-44. The anode material was a proprietary core-shell CGP-A12 ellipsoid particle (ConocoPhillips, USA). They are shown in Figure S4
in SI and are, respectively, labeled as “MCMB,” “SFG,” and “CGP” in the text.
To oxidize the materials, carbon powders were placed in alumina boats, heated in alumina tube furnaces (open to air) to 570 °C, and held at that temperature for 4 h. To hydrogenate graphite, we took oxidized graphite particles, placed them into quartz tube furnaces (under flowing Ar gas), and heated them at 800 °C under flowing H2
gas (at 0.5 L·min−1
) for 8 h [24
Micromeretics Tristar II porosimeter used N2
adsorption to quantify the specific surface area (SSA) of the functionalized materials. Samples were outgassed at 105 C for 24 h, and N2
sorption was carried out in the 0.05–0.995 P·P0−1
range using a liquid nitrogen bath (−193 °C). Brunauer-Emmett-Teller (BET) SSA was calculated in the 0.05–0.30 P·P0−1
] using Tristar II 3020 V1.03 Software (Micromeretics Instrument Corporation, Norcross, GA, USA).
X-Ray Photoelectron Spectroscopy (XPS) identified the surface chemistry and provided elemental analysis and functional group content of the graphite materials. Physical Electronic VersaProbe 5000 (VersaProbe, Chanhassen, MN, USA), with a 100 µm monochromatic Al-Kα X-ray beam Kratos AXIS surface analysis system (Ultra DLD), was used for performing the XPS measurements. CasaXPS Version 2.3.16 RP 1.6 software (Casa Software Ltd, Teignmouth, UK) deconvoluted the functional group peaks. Additionally, thermogravimetric analysis (TGA) with a TGA7 Perkin-Elmer instrument (Perkin-Elmer Corporation, Waltham, MA, USA) quantified total functional groups on each material. Samples were placed in Pt crucibles, heated to 750 °C at 5 °C·min−1 in a N2 environment, and held at that temperature for 1 h. Prior to temperature ramp-up, samples were held at 110 °C for 2 h in an inert atmosphere to desorb water or other chemical species. PYRIS Software (Perkin-Elmer) analyzed results.
Carbon structure of materials was evaluated using a Rigaku Ultima III X-Ray Diffractometer (XRD). This method evaluated both initial and electrochemically cycled materials. The instrument used 20 kV accelerating voltage in Bragg-Brentano mode. Materials were analyzed in the 5–50° 2θ range at a 2° 2θ min−1 scan rate. MDI Jade 7 Software (Materials Data, Inc., Livermore, CA, USA) applied baseline corrections and conducted peak fitting.
An amount of 2.0 M [Li+
] in monofluoroethylene carbonate (FEC)/ethylmethyl carbonate (EMC) 1:1(v
) mixture, along with a tris(hexafluoro-iso-propyl)phosphate (HFIP) additive, acted as the electrolyte for all cells (schematics of all electrolyte components shown in Figure S5
in SI). Graphite powders were mixed with poly(vinylidene) fluoride (PVDF) binder and conductive carbon black in n-methyl-2-pyrrolidone (NMP) solvent in an 80:10:10 ratio for electrode assembly. Cathode slurries were applied on a thin coating onto carbon-coated aluminum current collectors using a 30 μm doctor blade; anode materials were applied onto copper foil using a 10 µm doctor blade. The resulting mass loading of electrodes was 3.5 mg·cm−2
. Hoshen 2035 coin cells (1.59 cm diameter) were used to test both half-cell and full-cell configurations. Half-cells used Li metal counter electrode disks. Two layers of Celgard 2400 separator, along with a sheet of non-woven polypropylene fiber mat were used as separators. Cathode test cans were aluminum clad/Ni plated/stainless steel with a Pt tab ultrasonically welded inside the can to make direct contact with the coated electrode, while anode cans were composed of 316 stainless steel (assembly shown in Figure S6
in SI). Cathode: anode electrodes maintained a 3:1 mass ratio in full dual-intercalation cells.
Electrochemical performance was measured using a MACCOR Series 4000 battery tester (MACCOR Corporation, Tulsa, OK, USA). Cathode half-cells were charged/discharged using a C/10 rate (assuming a theoretical 90 mAh/g intercalation capacity, based on active material mass) using galvanostatic cycling. Starting at Open Circuit Potential (OCP), cells were preconditioned using single cycles in the 4.0–4.2 V, 4.0–4.6 V, 4.0–4.8 V, and 4.0–5.0 V range (vs. Li/Li+). Subsequently, cells were cycled between 4.0 and 5.2 V up to 200 cycles. Separately, cells were cycled at a C/2 rate with no preconditioning. Anode cells were cycled at C/10 rate (based on a 300 mAh/g expected intercalation capacity) between 0.5 V and 35 mV vs. Li/Li+. Dual intercalation cells were tested using the same pre-conditioning procedures as cathode half cells, with a vertex potential of 5.1 V. Polarization potentials were calculated from the inflection point in the voltage vs. capacitance relationship of the 1st full voltage range cycle.
Electrochemical Impedance Spectroscopy (EIS) measured the charge transfer resistance between electrodes and current collectors of coin cells before and after cycling with a Solartron SI 1287 potentiostat. Voltage oscillated with ±10 mV amplitude (vs. OCP) at a dampening frequency (106
Hz). Charge transfer resistance (RT
) was calculated by the Re (Z) size of the semi-circular region at mid-range oscillation frequencies [26
We were able to evaluate the influence of hydrogen- and oxygen-containing surface functional groups on efficiency and charge storage densities of graphite cathodes and anodes in dual-intercalation batteries. Particles with flake morphologies featured more exposed edge sites and were more susceptible to chemical treatments. While oxygen functional groups improved cyclability of cathodes, hydrogen functional groups were most beneficial for anode materials. Conversely, hydrogenated cathodes and oxidized anodes were most detrimental to electrochemical performance. Therefore, influence of specific functional groups strongly depended on the applied potential and matched with specific intercalated ions. When hydrogenated anodes were coupled with oxidized cathodes in all-graphite dual-intercalation cells, the resulting batteries demonstrated higher capacities, cyclabilities, and coulombic efficiencies than their untreated counterparts.
We relied on environmentally benign and economically affordable surface treatments to improve the performance of these all-graphite batteries and, subsequently, confirmed their status as a cost-effective solution for microgrid energy storage. We provided additional fundamental insight into the dual-intercalation process, and, in particular, the effects of anion insertion into graphite. Multiple other surface treatments, such as amination and moderate fluorination, remain promising pathways to even greater efficiencies and long-term stable cyclabilities; they must be explored in subsequent efforts. Finally, while surface chemistry-dependent differences suggested SEI formation on cathodes during anion intercalation, the structure, composition, and the mechanism of formation of this layer on positively charged electrodes remains unclear. Ongoing investigation into these processes will have broad implications for dual-intercalation batteries and apply to both graphite electrodes and other emerging materials, such as layered aluminum, transition metal carbide carbides, and boron nitrides.