Spencer A. Langevin,
Courtney A. McHale
,
Tanner Hamann
and
Jesse S. Ko
*
Research and Exploratory Development Department, Johns Hopkins University Applied Physics Laboratory, Laurel, MD 20723, USA. E-mail: Jesse.Ko@jhuapl.edu
First published on 30th July 2025
The development of advanced anode materials is critical for enhancing the energy density and safety of lithium-ion batteries (LIBs). Here, we demonstrate a blended anode comprising lithium titanate (LTO) and tin–antimony (SnSb) alloy, investigating the trade-offs between capacity and cycle life across varying LTO:
SnSb ratios. The near-zero strain properties of LTO provide mechanical durability by mitigating the volume expansion issues of SnSb when present at levels <20 wt% in the blended anode. This system delivers a favorable half-cell potential (∼0.5–1.5 V vs. Li/Li+) and specific capacities ranging from 300 to 500 mAh g−1.
Presently, Si anodes dominate anode materials research, owing to its high theoretical capacity (3580 mAh g−1 for Li15Si4 at room temperature), natural abundance, and low half-cell potential (∼0.37 V vs. Li/Li+).26–29 However, the large volume expansion of Si (300%) and poor electrical conductivity has led to challenges in commercialization.26–29 To compensate for these shortcomings, the concept of blending Si with gr (Si/gr) as a composite anode has gained significant traction in an attempt to combine the advantageous properties of gr (high electrical conductivity and structural stability) with Si (high theoretical capacity).30–34 This composite anode is designed to allow gr to act as a mechanically-stable matrix and provide a porous network that accommodates Si volume expansion, successfully improving cycle life.29,30,32–34 Though Si/gr has merits in terms of energy density, the low half-cell potentials still pose a safety concern due to the potential for lithium dendrite growth.
As an alternative to Si/gr anode composites, we propose an alternative selection of blended anode materials—such as combining oxide-based anodes and Sn- and Sb-based alloying anodes5,35,36—as a unique approach to bridging a critical gap in materials selection (Fig. 1; 0.5–1.5 V vs. Li/Li+ and 350–600 mAh g−1).37 Extensive research has been conducted on titanate and niobate anode materials, which have demonstrated high-rate capability and enhanced safety; yet, their specific capacities are typically limited to <300 mAh g−1. Similarly, Sb, Sn, and SnSb intermetallic alloys have shown significant improvements over Si in terms of electrical conductivity, yet their cycle life remains a challenge.5 To achieve an optimum balance between half-cell potential (>0.5 V vs. Li/Li+) and specific capacity (∼300–500 mAh g−1), we demonstrate that systematically blending LTO with SnSb (LTO:
SnSb) results in an energy-dense anode material with improved cycle life. In addition, replacing the copper current collector for aluminum leads to lower cost and the potential for zero-volt storage. Given that LTO is a near-zero strain material,18,19,38 incorporating it into a conductive carbon matrix provides mechanical support for SnSb's volume expansion, thereby improving structural stability. Though prior studies have investigated composites comprising LTO with Si,39 or Sn,40,41 a rigorous in-depth assessment that balances capacity and cycle life has yet to be reported. Moreover, SnSb, which offers innately greater cycle life versus Sn or Sb alone,42 has not been previously incorporated as a composite anode with LTO.
While strategies such as nanoporosity,43–45 hierarchical structures,46–48 and nanowires49–52 have improved cycle life for alloying anodes, their complex synthesis routes present significant barriers to integration into current manufacturing techniques like roll-to-roll coating and calendaring. Moreover, Luo et al. critically evaluated the impact of inter- and intra-particle porosity in alloying anodes (e.g., Si), emphasizing that careful engineering of inter-particle porosity is required to achieve high gravimetric and volumetric energy densities without compromising cycle life.53 Beyond compositional optimization, the blending of LTO and SnSb in this work demonstrates a path toward practical performance while maintaining compatibility with standard lithium-ion battery manufacturing processes (e.g., coating and calendaring), providing a technologically-relevant implementation approach that can facilitate industry adoption with existing infrastructure.
The primary criterion for constructing high-capacity blended anodes was to first ensure phase purity of the selected alloying (SnSb) and intercalation (LTO) materials. Rietveld refinement of the X-ray diffraction patterns for SnSb and LTO confirm their phase purity, with all primary reflections closely matching their respective reference patterns (Fig. 2), as indicated by their low-profile residual goodness-of-fit factor (Rwp < 10%). To further characterize the materials, scanning electron microscopy (SEM) was performed to analyze the particle sizes of SnSb and LTO. The average particle size for SnSb was estimated to be 5–20 μm, while LTO exhibited primary particle sizes of ∼20 μm and secondary particle sizes of ∼0.5 μm. Of note, though this is beyond the scope of this work, we hypothesize that smaller nanoparticles incorporated with micron-sized LTO particles will further improve the structural rigidity of the composite anode by further minimizing inter-particle porosity.53 These two materials were then mixed in several mass ratios (wt:
wt), which are denoted as the following: (i) 100 wt% LTO (10
:
0); (ii) 90 wt% LTO and 10 wt% SnSb (9
:
1); (iii) 80 wt% LTO and 20 wt% SnSb (8
:
2); (iv) 50 wt% LTO and 50 wt% SnSb (5
:
5); (v) 30 wt% LTO and 70 wt% SnSb (3
:
7); and (vi) 100 wt% SnSb (0
:
10). Based on these ratios between LTO and SnSb (ranging from 1
:
9 to 9
:
1), the projected specific energies range between 147–282 Wh kg−1 (Fig. S1; projections were based on an areal capacity of 3.3 mAh cm−2 for the anode, 3.0 mAh cm−2 for the LiNi0.8Co0.15Al0.05O2 (NCA) cathode, and an electrolyte overfill of 60%, in a 5 Ah pouch cell form factor). These formulations were prepared as carbon-composite anodes utilizing active material, conductive carbon, and polymeric binder coated onto aluminum foil. After calendaring, the final electrode thickness for all coatings was measured to be ∼60 μm with a porosity of ∼40%.
Half-cell electrochemical measurements were conducted using lithium metal as both the reference and counter electrodes to evaluate the average half-cell potential and specific capacities of the blended anodes (Table 1). As expected, LTO (10:
0) exhibits an average discharge potential of 1.55 V vs. Li/Li+ with a C/10 specific capacity of 167.9 mAh g−1, while SnSb (0
:
10) displays a 0.64 V average potential and a capacity of 748.1 mAh g−1 (Fig. 3a). LTO demonstrates excellent high-rate capability, with only a 5% capacity loss after increasing to 2C (159.9 mAh g−1). In contrast, SnSb experiences a 29% capacity drop at to 2C (531.2 mAh g−1) due to the slower kinetics of the alloying reaction. With increasing amounts of SnSb in the LTO
:
SnSb blend, the charge–discharge voltage profiles (Fig. 3a, overlays in Fig. S2) reveal a gradual increase in blended anode capacity, accompanied by a commensurate decrease in average discharge potential. The 9
:
1 blend maintains a slightly lower discharge potential than LTO (1.14 V) while increasing total anode capacity to 252.5 mAh g−1 at C/10, with high capacity retention when cycling at 2C (222 mAh g−1). For the 8
:
2 blend, the discharge potential drops ∼200 mV (0.93 V), and capacity increases to 344.7 mAh g−1 at C/10, retaining 89% capacity at 2C (306.7 mAh g−1). The 7
:
3 ratio exhibits strong capacity retention at a 0.80 V average discharge potential, with 456.9 mAh g−1 at C/10 and 402.1 mAh g−1 at 2C. For the higher SnSb loadings (5
:
5 and 3
:
7), the average discharge potentials decrease to 0.73 V and 0.67 V, respectively, with specific capacities ranging from 552–689 mAh g−1, though rate performance decreases by over 15% when cycling at 2C. Rate capability overlays of specific capacity versus cycle number for this material series are shown in Fig. S3. Based on these half-cell performance metrics, we then fabricated full cells pairing these anodes with technologically-relevant NCA cathodes to assess practical cell voltage, rate capability, and device-level performance.
LTO![]() ![]() ![]() ![]() |
Average discharge potential (V vs. Li/Li+) | C/10 capacity (mAh g−1) | C/5 capacity (mAh g−1) | C/2 capacity (mAh g−1) | 1C capacity (mAh g−1) | 2C capacity (mAh g−1) |
---|---|---|---|---|---|---|
10![]() ![]() |
1.55 | 167.9 | 167.6 | 166.8 | 164.6 | 159.9 |
9![]() ![]() |
1.14 | 252.5 | 248.6 | 239.0 | 230.7 | 222.3 |
8![]() ![]() |
0.93 | 344.7 | 340.0 | 327.1 | 325.3 | 306.7 |
7![]() ![]() |
0.80 | 456.9 | 454.5 | 437.5 | 431.2 | 402.1 |
5![]() ![]() |
0.73 | 552.0 | 550.7 | 528.5 | 508.8 | 482.3 |
3![]() ![]() |
0.67 | 689.0 | 684.4 | 656.6 | 630.6 | 598.1 |
0![]() ![]() |
0.64 | 748.1 | 700.7 | 621.6 | 570.6 | 531.2 |
Full cells were tested within a voltage range of 1.0–2.8 V for LTO‖NCA (10:
0‖NCA) and 1.0–3.8 V for 9
:
1‖NCA, 8
:
2‖NCA, 7
:
3‖NCA, 5
:
5‖NCA, and 3
:
7‖NCA, optimizing both anode and cathode performance based on their respective half-cell potential windows. The potential window for the LTO
:
SnSb‖NCA cells were determined first by evaluating a range of upper charge voltages (3.6, 3.8, and 4.0 V; Fig. S4) using 7
:
3‖NCA as a representative cell, which showed that 1.0 to 3.8 V was ideal for maintaining good capacity retention. Subsequently, the N
:
P ratio was determined by evaluating two ratios (∼0.8 and ∼1.1), and ∼0.8 was selected, as this showed better capacity retention (Fig. S5). As shown in Fig. 3b (direct overlays are shown in Fig. S6), the 10
:
0‖NCA full cell exhibits a characteristic charge–discharge profile with an average discharge voltage of 2.11 V and a C/10 capacity of 180.5 mAh g−1 (Table 2). Unlike the near-constant capacities observed in half-cell testing, the 10
:
0‖NCA full cell shows a modest capacity drop at 2C (131 mAh g−1) due to the rate limitations of the NCA cathode (Fig. 3c). With increasing SnSb content, the average discharge voltage proportionally increases, accompanied by variations in discharge capacity (Fig. 3b and Table 2). The 9
:
1‖NCA full cell achieves a capacity of 250.6 mAh g−1 at C/10 with an average discharge voltage of 2.43 V, retaining 193.4 mAh g−1 at 2C. The 8
:
2‖NCA full cell achieves 296.1 mAh g−1 at C/10, maintaining 230.7 mAh g−1 at 2C. The 7
:
3‖NCA full cell follows this trend with improved capacity and rate retention. For higher SnSb loadings, performance trends begin to shift, and the 5
:
5‖NCA full cell exhibits a high C/10 capacity of 485.9 mAh g−1, but capacity retention drops significantly at 2C (384.6 mAh g−1). The 3
:
7‖NCA full cell achieves the highest C/10 capacity (637.3 mAh g−1) and 502.6 mAh g−1 at 2C; however, it also indicates the greatest capacity fade when cycled at C/2 for 100 cycles (Fig. 3d). For full cells using LTO
:
SnSb blended anodes with medium-to-high SnSb content (i.e., 5
:
5 to 8
:
2), long-term cycling stability at C/2 collapses at ∼60 cycles. The 8
:
2‖NCA full cell exhibits only modest degradation at 60 cycles, suggesting this ratio was the upper limit for SnSb incorporation to balance cycle life with discharge capacity. In contrast, both the 9
:
1‖NCA and 10
:
0‖NCA full cells maintains excellent capacity retention over 100 cycles at C/2, attributed to reduced volume expansion and pulverization of SnSb. Based on the average discharge voltage and C/10 capacities for this series of LTO
:
SnSb full cells, specific energy metrics were calculated normalizing to both the total anode and cathode mass (not including mass of current collectors, electrolyte and packaging). These estimates show that achievable specific energy values range between 213.5 to 453.3 Wh kg−1 at the materials level; yet, we note that ∼60% of this value should be taken as a closer approximation towards technological relevance (see also Fig. S1 on projections that incorporate the mass of all cell components, packaging and electrolyte fill).
LTO![]() ![]() ![]() ![]() |
Average discharge voltage (V) | C/10 capacitya (mAh g−1) | C/5 capacitya (mAh g−1) | C/2 capacitya (mAh g−1) | 1C capacitya (mAh g−1) | 2C capacitya (mAh g−1) | Specific energyb (Wh kg−1) |
---|---|---|---|---|---|---|---|
a Capacities normalized by the anode active material mass.b Specific energy normalized to both total anode and cathode mass (including carbon and binder), not including current collector, electrolyte, and packaging. | |||||||
10![]() ![]() |
2.11 | 180.5 | 172.9 | 158.6 | 146.5 | 131.0 | 213.5 |
9![]() ![]() |
2.43 | 250.6 | 242.9 | 225.9 | 210.9 | 193.4 | 294.2 |
8![]() ![]() |
2.62 | 296.1 | 287 | 268.8 | 251.3 | 230.7 | 351.4 |
7![]() ![]() |
2.73 | 387.7 | 371.9 | 345.6 | 318.0 | 284.3 | 397.1 |
5![]() ![]() |
2.79 | 485.9 | 474.8 | 451.6 | 421.9 | 384.6 | 424.8 |
3![]() ![]() |
2.83 | 637.3 | 624.7 | 592.3 | 554.8 | 502.6 | 453.3 |
To investigate the mechanisms behind capacity fade in these blended anode materials, dQ/dV analyses were performed to track changes in peak intensities and voltage locations, which indicate electrochemical reactions leading to degradation.54–57 By converting the charge–discharge profiles into dQ/dV plots at cycle 5 and 100 for this series of blended anode materials, distinct trends emerge based on the LTO:
SnSb ratio. For the 10
:
0‖NCA full cell, the dQ/dV peak intensity and broadness remain unchanged between cycles 5 and 100, indicating stable performance with no signs of LTO anode degradation (Fig. S7). In contrast, the 9
:
1‖NCA full cell exhibits two distinct peaks (∼2.1 V for LTO and ∼3.0 V for SnSb) at cycle 5. After 100 cycles, the higher-voltage SnSb peaks diminish while the LTO peak remains stable, suggesting the primary degradation mechanism in the LTO
:
SnSb blended anode is alloying-related degradation of the SnSb. For the 8
:
2‖NCA cell, a similar trend is observed but with an additional shift in the LTO peak (∼2.1 V → ∼2.2 V), indicating significant aging of LTO due to SnSb-induced volume expansion and pulverization. This shift in the voltage for LTO-based cells corroborates prior reports that investigated aging behavior under various scenarios, particularly under high-rate conditions (i.e. >1C), and their dQ/dV analyses show that the surface of the anode showed signs of fracture.58,59 This effect becomes more apparent in full cells comprising >30% SnSb (Fig. S7), where the higher-voltage SnSb peak nearly disappears after 100 cycles, leaving behind two peaks at ∼2.5 V, signaling a significant electrochemical transformation of the blended anode. For the 3
:
7‖NCA full cell, severe degradation leads to a near complete loss of blended anode capacity, with no distinguishable dQ/dV peaks after 100 cycles. This further confirms the detrimental impact of excessive SnSb content on long-term cycle stability of the LTO
:
SnSb blended anode.
We hypothesize that the capacity degradation of the LTO:
SnSb blended anodes with >30 wt% SnSb was primarily due to the inherent volume expansion and pulverization of SnSb, while the lack of degradation for LTO
:
SnSb blended anodes with <20 wt% SnSb may be attributed to the presence of a more compact LTO network surrounding the SnSb material, thus potentially mitigating significant pulverization, which is the subject of future studies. To investigate this, SEM, coupled with energy dispersive spectroscopy (EDS) for chemical mapping of the anodes (Fig. 4), was conducted on pre- and post-cycled samples of 9
:
1, 8
:
2, and 5
:
5 LTO
:
SnSb blended anodes after 100 cycles at C/2. Additional SEM/EDS micrographs and EDS elemental maps for all blended anodes are available in the SI (Fig. S8–S14). As expected, for the 10
:
0 anode (100% LTO), SEM images (Fig. S8) reveal a compact, dense anode layer. In contrast, SnSb naturally forms a rougher, more porous network (Fig. S9). For the 9
:
1 anode, the Ti EDS map shows a near-continuous film on the electrode surface before cycling; after 100 cycles, minimal surface changes are observed and the visible SnSb particles retain their original size. However, in the 8
:
2 anode, two distinct changes emerge: (i) before cycling, the LTO network appears well-structured, but after 100 cycles, a larger LTO particle network develops; and (ii) the SnSb particles pulverizes, resulting in smaller aggregates of the original SnSb powder after 100 cycles. This effect is more pronounced in the 5
:
5 anode, where the LTO network becomes significantly disrupted and there is a greater presence of pulverized SnSb particles after 100 cycles. This behavior aligns with observations from the 0
:
10 anode (100% SnSb), where extensive SnSb pulverization is clearly visible after cycling (Fig. S9). This pulverization behavior of alloying anodes (e.g., Sn, SnSb, and Sb) is consistent with prior investigations that probed deeper into understanding the effects of performance degradation in SnSb,60 and mitigations therein with nanoporous forms of Sn and SnSb.43,44 These findings suggest that limiting the SnSb content <20 wt% is crucial for long-term LTO
:
SnSb blended anode stability, consistent with the dQ/dV analyses. Notably, the combination of the dQ/dV analyses and SEM/EDS provides only cursory insight, and in future studies, techniques such as operando X-ray micro-computed tomography and transmission X-ray microscopy can reveal a more quantitative interpretation of this degradation behavior in composite anodes with high spatial resolution.
In this study, we developed and demonstrated a blended anode concept that combines zero-strain LTO, an oxide-based anode material, with high-capacity SnSb, which operates via alloying. By systematically varying the LTO-to-SnSb mass ratios, we found that maintaining the SnSb content below 20 wt% enables high specific capacities (250–350 mAh g−1) at average full-cell discharge voltages of 2.43–2.62 V, while achieving excellent cycle retention (>80% after 100 cycles at C/2 rates). This strategy highlights a technologically-relevant pathway for seamless integration into existing lithium-ion battery manufacturing processes. Future research focusing on carbon electrode architecture design, such as high-areal-capacity carbon,61–63 can further extend the cycle life of alloying anodes, advancing the practical viability of this emerging blended anode concept.
Experimental details; estimated gravimetric and volumetric energy densities; additional half-cell charge–discharge voltage profiles; additional half-cell rate capability data; upper voltage limit cutoff; N:
P ratio optimization; additional full-cell charge–discharge voltage profiles; dQ/dV analyses; additional scanning electron micrographs and energy dispersive spectroscopy elemental maps. See DOI: https://doi.org/10.1039/d5ta05287d.
This journal is © The Royal Society of Chemistry 2025 |