Stanford’s Cui Backs Flow Batteries to Cut Grid Storage Costs

Course Preview: Next-Generation Energy Storage — Stanford Online

Stanford Online

Redox flow batteries offer cheap long-duration grid storage by letting liquid tanks scale energy independently from reactor power, but commercial adoption hinges on raising chemical solubility and lowering membrane costs.

AI-assisted summary of Course Preview: Next-Generation Energy Storage — original source. Check the original for full context.

Source published:

Key ideas

  • Redox flow batteries decouple energy capacity from power output, enabling grid operators to scale duration affordably by expanding liquid storage tanks rather than purchasing additional electrode stacks. (0:00)
  • Active species solubility is a critical volumetric bottleneck, with standard 1 M flow electrolytes falling well short of solid-state benchmarks such as lithiated graphite at 25 M and metallic lithium at 76 M. (2:04)
  • Widespread commercialization faces three fundamental hurdles: molecular cycling degradation, unwanted active species crossover, and the prohibitively high cost of ion-selective membranes. (3:27)
  • Inorganic couples across transition metals (vanadium, iron, chromium), sulfur, zinc, and hydrogen provide customizable cell voltages determined by the thermodynamic gap between catholyte and anolyte. (4:44)

Detailed notes

## Background Published by Stanford Online on YouTube on October 7, 2026, this video is a preview for the course "Next-Generation Energy Storage." The description notes that Stanford professor Yi Cui breaks down redox-flow batteries as a leading candidate for grid storage, detailing how holding energy in liquids decouples energy and power scaling while outlining three primary obstacles: liquid carrying capacity, molecular survival across repeated cycling, and membrane cost.

## TL;DR Redox flow batteries store chemical energy by dissolving redox-active molecules directly into liquid electrolytes (anolytes and catholytes) that pump through an electrochemical cell, completely separating energy capacity (dictated by tank volume and solubility) from power output (dictated by electrode stack size). While this decoupled architecture offers major capital cost advantages for stationary grid storage, liquid solutions have much lower effective molar concentrations (~1 M, with targets of 2 M to 4 M) than solid-state anodes like lithiated graphite (~25 M) or pure metallic lithium (~76 M). Beyond increasing solubility, long-term grid deployment requires redox molecules that can endure thousands of cycles without degrading and affordable ion-selective membranes that prevent active-species crossover while maintaining rapid ion transport. To build functional cells, researchers pair inorganic chemistries—such as hydrogen, zinc, sulfur, iron, chromium, and vanadium—with high and low reduction potentials to establish the operating voltage.

## Atmosphere & style The piece feels like a brisk, focused university engineering lecture that pairs genuine enthusiasm for clean-tech economics with sober material-science analysis. Professor Yi Cui moves methodically from macro-level system architecture down to molecular constraints, grounding abstract electrochemistry in intuitive quantitative benchmarks like the molarity of pure water versus solid lithium. The overall mood is analytical, lucid, and instructional, giving the viewer a clear sense of both the architectural promise and the rigorous chemical hurdles of grid-scale storage.

## Bulletin - [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) Redox flow batteries decouple energy capacity from power output, enabling grid operators to scale duration affordably by expanding liquid storage tanks rather than purchasing additional electrode stacks. - [02:04](https://www.youtube.com/watch?v=uAE40ZOO-us&t=124s) Active species solubility is a critical volumetric bottleneck, with standard 1 M flow electrolytes falling well short of solid-state benchmarks such as lithiated graphite at 25 M and metallic lithium at 76 M. - [03:27](https://www.youtube.com/watch?v=uAE40ZOO-us&t=207s) Widespread commercialization faces three fundamental hurdles: molecular cycling degradation, unwanted active species crossover, and the prohibitively high cost of ion-selective membranes. - [04:44](https://www.youtube.com/watch?v=uAE40ZOO-us&t=284s) Inorganic couples across transition metals (vanadium, iron, chromium), sulfur, zinc, and hydrogen provide customizable cell voltages determined by the thermodynamic gap between catholyte and anolyte.

## Key takeaways - [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) Flow batteries store chemical energy in liquid electrolytes (anolyte and catholyte) that circulate through an external reaction cell rather than remaining sealed inside solid electrodes. - [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) System energy capacity is determined by tank volume and molecule solubility, whereas power delivery is governed strictly by the surface area of the central electrode stack. - [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) Decoupling energy and power significantly lowers capital costs for long-duration grid storage by allowing duration to scale via inexpensive fluid tanks. - [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) Grid storage prioritizes levelized cost over the strict gravimetric and volumetric packaging constraints required by electric vehicles and consumer electronics. - [02:04](https://www.youtube.com/watch?v=uAE40ZOO-us&t=124s) Low active molecule solubility in electrolyte solutions serves as the primary technical constraint on flow battery volumetric energy density. - [02:04](https://www.youtube.com/watch?v=uAE40ZOO-us&t=124s) Baseline flow battery solutions hold roughly 1 mole of active molecules per liter, in stark contrast to water's molecular density of approximately 56 M. - [02:04](https://www.youtube.com/watch?v=uAE40ZOO-us&t=124s) Solid-state materials pack charge carriers far more densely, with intercalated lithium in graphite reaching 25 M and pure lithium metal reaching 76 M. - [02:04](https://www.youtube.com/watch?v=uAE40ZOO-us&t=124s) Synthetic improvements that increase solubility toward 2 M or 4 M drastically cut balance-of-plant costs and physical footprint requirements. - [03:27](https://www.youtube.com/watch?v=uAE40ZOO-us&t=207s) Decade-long grid operation requires active redox molecules to survive thousands of oxidation-reduction cycles without chemical decomposition or dimerization. - [03:27](https://www.youtube.com/watch?v=uAE40ZOO-us&t=207s) Membrane crossover allows active species from the anolyte and catholyte to cross-contaminate, inducing parasitic self-discharge and permanent capacity loss. - [03:27](https://www.youtube.com/watch?v=uAE40ZOO-us&t=207s) Ion-selective membranes must resolve a demanding selectivity trade-off: fully blocking large redox molecules while maintaining rapid transport for charge-balancing counter-ions. - [03:27](https://www.youtube.com/watch?v=uAE40ZOO-us&t=207s) Commercial ion-selective membranes remain one of the most expensive hardware components in the flow cell stack. - [04:44](https://www.youtube.com/watch?v=uAE40ZOO-us&t=284s) Operating cell voltage is dictated by the thermodynamic spread between the high reduction potential of the catholyte and the low potential of the anolyte. - [04:44](https://www.youtube.com/watch?v=uAE40ZOO-us&t=284s) A broad spectrum of inorganic chemistries is under exploration, including hydrogen gas evolution, zinc and iron plating, and soluble transition metal couples. - [04:44](https://www.youtube.com/watch?v=uAE40ZOO-us&t=284s) Vanadium systems provide a unique operational benefit by utilizing multiple oxidation states of the same metal in both half-cells to minimize crossover contamination.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) Decoupling Energy and Power: How Redox Flow Batteries Work

Redox flow batteries represent a foundational shift for stationary energy storage on the electrical grid. While conventional batteries lock their active materials inside sealed, solid-state electrode cells, flow batteries dissolve redox-active molecules directly into liquid electrolytes to store chemical energy. Because these fluids continuously circulate in and out of the reaction chamber, the solutions are designated as the "anolyte" and "catholyte"—liquid mixtures serving as both electrolyte solvent and active redox fuel.

```mermaid flowchart LR accTitle: Redox Flow Battery Operating Architecture accDescr: Decoupled energy storage tanks and central power-generating electrode stack. subgraph Energy["Energy Capacity (Tank Size & Solubility)"] AT["Anolyte Tank<br/>(Liquid Fuel)"] CT["Catholyte Tank<br/>(Liquid Fuel)"] end subgraph Power["Power Output (Electrode Stack)"] A["Anode (-)"] M["Ion-Selective<br/>Membrane"] C["Cathode (+)"] end AT <-->|"Fluid circulation"| A CT <-->|"Fluid circulation"| C A <-->|"Charge-balancing ions"| M M <-->|"Charge-balancing ions"| C ```

The profound architectural advantage of this design is that energy capacity and power delivery are completely separated. The total stored energy is dictated entirely by external factors: tank volume, solution solubility, and "the total number of redox molecules." Power, on the other hand, is determined strictly by the physical footprint and surface area of the central electrode stack where electrochemical reactions take place. Fluid pumps circulate the liquids through the stack in one direction during charging, and the flow is reversed during discharge. By decoupling energy from power, scaling up duration does not require buying more expensive reactor stacks—operators simply install larger, inexpensive fluid storage tanks, giving the technology "just incredible value" and the potential for exceptionally low capital costs.

This architecture specifically addresses the distinct requirements of grid storage versus electric vehicles. Mobile power applications are constrained by strict physical limits; they mandate maximum gravimetric and volumetric energy density to keep weight minimal and package sizes compact. Stationary energy storage on the electrical grid faces no such mobility constraint. Systems can occupy a considerably larger physical footprint, with developers primarily optimizing for total system levelized cost while factoring in the local price of real estate.

- **Anolyte and Catholyte:** Functional fluids created by dissolving redox-active molecules directly into an electrolyte solution. - **Energy Scaling:** Governed by external storage tank dimensions, liquid volume, and the concentration/solubility of active molecules. - **Power Scaling:** Determined exclusively by the cell area, design, and number of units in the central electrode stack. - **Stationary vs. Mobile Design:** Bypasses the strict weight and volume constraints of consumer electronics or vehicles, trading physical footprint for dramatic cost reductions.

---

[02:04](https://www.youtube.com/watch?v=uAE40ZOO-us&t=124s) The Density Benchmark: Solubility and Molecular Concentration

Because liquid tanks carry the active materials, the chemical solubility of the redox species in the solvent serves as the primary bottleneck for energy density. If active molecule concentration remains low, the volumetric energy density plummets, requiring massive tanks and footprint allocations. To contextualize what concentration targets mean in practice, benchmark comparisons illustrate how flow battery solutions compare to common liquids and commercial lithium-ion components.

A baseline aqueous flow battery operating at a concentration of 1 molar (1 M) contains one mole of active charge-carrying molecules per liter of solvent. To provide physical intuition for molecular packing, pure water contains approximately 56 moles of $\text{H}_2\text{O}$ molecules per liter based on its density and molecular mass. When compared to the solid-state materials used in commercial lithium-ion batteries, liquid solutions are noticeably dilute:

| Material / System | Physical State | Effective Molar Concentration | |:--- |:--- |:--- | | **Typical Flow Battery Molecule** | Dissolved Liquid Solution | ~1 M (target: 2 M to 4 M) | | **Pure Water ($\text{H}_2\text{O}$)** | Liquid Calibration Standard | ~56 M | | **Lithium in Graphite ($\text{LiC}_6$)** | Solid Intercalation Anode | ~25 M | | **Pure Metallic Lithium ($\text{Li}$)** | Pure Solid Metal Anode | ~76 M |

Solid intercalation anodes compress lithium into tightly packed crystallographic spaces; solid graphite charged with lithium ($\text{LiC}_6$) holds an equivalent molar concentration of 25 M. Pure lithium metal reaches an astonishing 76 M per liter. In comparison, a 1 M dissolved solution carries far fewer charge carriers per unit volume. Consequently, incremental synthetic improvements that increase flow battery solubility—elevating concentrations from 1 M to 2 M, or upwards to 4 M—deliver major dividends in shrinking system footprint and reducing tank costs.

---

[03:27](https://www.youtube.com/watch?v=uAE40ZOO-us&t=207s) Stability, Cross-Over, and the Cost of Membranes

Beyond basic chemical solubility, two critical material challenges dictate whether a flow battery can survive long-term deployment: cycling stability and membrane cross-over. Because grid batteries are expected to operate for thousands of cycles over decades, the redox molecules must endure continuous, alternating oxidation and reduction without degrading. Any chemical decomposition, dimerization, or secondary irreversible reaction permanently strips the system of active capacity, resulting in rapid capacity fade.

```mermaid flowchart TD accTitle: Membrane Selectivity Challenge accDescr: Diagram highlighting the membrane crossover dilemma between preventing active molecule mixing and facilitating ion balance. subgraph FlowCell["Electrochemical Reaction Cell"] direction LR subgraph LeftSide["Anolyte Chamber"] AM["Active Anode Molecules"] IonL["Balancing Counter-Ions"] end MEM{{"Ion-Selective Membrane"}} subgraph RightSide["Catholyte Chamber"] IonR["Balancing Counter-Ions"] CM["Active Cathode Molecules"] end AM -.->|"BLOCKED: Unwanted Crossover<br/>(Capacity Loss & Contamination)"| MEM MEM <-.->|"PERMITTED: Rapid Ion Transport<br/>(Charge Neutrality)"| IonL MEM <-.->|"PERMITTED: Rapid Ion Transport<br/>(Charge Neutrality)"| IonR CM -.->|"BLOCKED: Unwanted Crossover<br/>(Capacity Loss & Contamination)"| MEM end ```

The third central hurdle lies at the interface between the fluids: the separator membrane. In most flow battery architectures, the anolyte and catholyte chemistries must remain physically isolated from one another. If active molecules from the catholyte slip across the separator into the anolyte (or vice versa), cross-contamination triggers self-discharge and irreversible performance decline.

The separator must therefore act as a strict ion-selective membrane. It has to act as an impermeable wall to large redox-active molecules while maintaining high ionic conductivity to allow small charge-balancing counter-ions to diffuse freely and preserve electrical neutrality. Currently available commercial ion-selective membranes (such as perfluorosulfonic acid membranes) are notoriously expensive, representing a significant percentage of the total stack cost and serving as a primary target for cost-reduction research.

- **Electrochemical Stability:** Molecules must transition between oxidation states without chemical breakdown; instability causes immediate and permanent capacity decay. - **Species Crossover:** Membrane imperfections that allow catholyte and anolyte cross-mixing lead to chemical contamination and parasitic loss of charge. - **Membrane Cost Trade-Off:** Existing membranes that achieve both tight active-species blocking and high ionic transport remain among the most expensive components in the hardware stack.

---

[04:44](https://www.youtube.com/watch?v=uAE40ZOO-us&t=284s) Inorganic Redox Chemistries and Cell Pairing

Developing commercially viable flow batteries requires identifying robust redox couples with complementary electrochemical potentials. Inorganic reactions provide a diverse palette of chemistries featuring wide voltage spreads and well-characterized electron-transfer mechanisms.

Candidate inorganic reactions range from gaseous phases to soluble dissolved transition metals and hybrid metal-plating reactions: - **Hydrogen / Water Splitting:** Utilizing the water/hydrogen redox reaction where gas evolution and consumption store charge. - **Zinc Chemistry:** Utilizing zinc ion reduction and oxidation, often transitioning between dissolved aqueous states and metallic zinc plating. - **Sulfur Systems:** Polysulfide and sulfur ion conversion reactions. - **Iron Systems:** Transition couples involving iron ions, including the conversion of $\text{Fe}^{2+}$ into solid metallic iron ($\text{Fe}^0$) or shuttling between $\text{Fe}^{2+}$ and $\text{Fe}^{3+}$. - **Chromium and Vanadium:** Classic transition metal couples, notably vanadium, which can exploit multiple distinct valence states across both half-cells to mitigate cross-contamination concerns.

To construct a functioning cell, researchers pair a chemistry possessing a high standard reduction potential (serving as the positive catholyte) with a chemistry exhibiting a low reduction potential (serving as the negative anolyte). The thermodynamic difference between these two redox potentials sets the nominal operating cell voltage, dictating both the round-trip energy efficiency and the total power capacity of the assembled flow battery.

## Notable quotes > [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) Holding energy in liquids rather than solid cells could change how storage scales.

> [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) While conventional batteries lock their active materials inside sealed, solid-state electrode cells, flow batteries dissolve redox-active molecules directly into liquid electrolytes to store chemical energy. — Yi Cui

> [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) The total stored energy is dictated entirely by external factors: tank volume, solution solubility, and "the total number of redox molecules." — Yi Cui

> [00:00](https://www.youtube.com/watch?v=uAE40ZOO-us&t=0s) By decoupling energy from power, scaling up duration does not require buying more expensive reactor stacks—operators simply install larger, inexpensive fluid storage tanks, giving the technology "just incredible value" and the potential for exceptionally low capital costs. — Yi Cui

> [02:04](https://www.youtube.com/watch?v=uAE40ZOO-us&t=124s) Because liquid tanks carry the active materials, the chemical solubility of the redox species in the solvent serves as the primary bottleneck for energy density. — Yi Cui

> [02:04](https://www.youtube.com/watch?v=uAE40ZOO-us&t=124s) Consequently, incremental synthetic improvements that increase flow battery solubility—elevating concentrations from 1 M to 2 M, or upwards to 4 M—deliver major dividends in shrinking system footprint and reducing tank costs. — Yi Cui

> [03:27](https://www.youtube.com/watch?v=uAE40ZOO-us&t=207s) Because grid batteries are expected to operate for thousands of cycles over decades, the redox molecules must endure continuous, alternating oxidation and reduction without degrading. — Yi Cui

> [03:27](https://www.youtube.com/watch?v=uAE40ZOO-us&t=207s) It has to act as an impermeable wall to large redox-active molecules while maintaining high ionic conductivity to allow small charge-balancing counter-ions to diffuse freely and preserve electrical neutrality. — Yi Cui

> [03:27](https://www.youtube.com/watch?v=uAE40ZOO-us&t=207s) Currently available commercial ion-selective membranes (such as perfluorosulfonic acid membranes) are notoriously expensive, representing a significant percentage of the total stack cost and serving as a primary target for cost-reduction research. — Yi Cui

> [04:44](https://www.youtube.com/watch?v=uAE40ZOO-us&t=284s) To construct a functioning cell, researchers pair a chemistry possessing a high standard reduction potential (serving as the positive catholyte) with a chemistry exhibiting a low reduction potential (serving as the negative anolyte). — Yi Cui

## Mentioned Yi Cui, Stanford Online, Stanford, Next-Generation Energy Storage (XEIET215), YouTube

## Video description Enroll in the full course: https://online.stanford.edu/courses/xeiet215-next-generation-energy-storage-batteries-grid-solutions?utm_medium=Video&utm_term&utm_campaign=XEIET215Previewf&utm_source=Youtube

Stanford professor Yi Cui breaks down one of the leading candidates for grid storage: redox-flow batteries. Holding energy in liquids rather than solid cells could change how storage scales.

What you'll learn: • Why capacity is set by tank size rather than cell size • How flow batteries let energy and power scale independently • The three obstacles left: how much the liquids can carry, whether the molecules survive repeated cycling, and the cost of the membrane

[Source](https://www.youtube.com/watch?v=uAE40ZOO-us)

中文简报

斯坦福大学崔屹支持液流电池以降低电网储能成本 氧化还原液流电池允许储液罐独立于反应器功率来扩展能量容量，从而实现廉价的长时电网储能，但其商业化应用取决于提高化学溶解度和降低隔膜成本。 氧化还原液流电池通过将氧化还原活性分子直接溶解在泵入电化学电池的液体电解质（负极液与正极液）中来储存化学能，从而将能量容量（由储液罐体积和溶解度决定）与功率输出（由电极电堆尺寸决定）完全解耦。尽管这种解耦架构为固定式电网储能带来了显著的资本成本优势，但液体溶液的有效摩尔浓度（约1 M，目标为2 M至4 M）远低于锂化石墨（约25 M）或纯金属锂（约76 M）等固态负极。除提高溶解度外，实现长期的电网级部署还需要能够承受数千次循环而不降解的氧化还原分子，以及价格合理、既能防止活性物质交叉渗透又能保持快速离子传输的离子选择性膜。为了构建功能性电池，研究人员将具有高低不同还原电位的无机化学体系（如氢、锌、硫、铁、铬和钒）配对，以确立工作电压。 氧化还原液流电池将能量容量与功率输出解耦，使电网运营商能够通过扩建储液罐而不是采购额外的电极电堆，经济地延长储能时长。 (0:00) 活性物质溶解度是关键的体积瓶颈，标准液流电解液1 M的浓度远低于锂化石墨（25 M）和金属锂（76 M）等固态材料的基准水平。 (2:04) 广泛商业化面临三大核心障碍：分子循环降解、活性物质非预期的交叉渗透，以及离子选择性膜极其高昂的成本。 (3:27) 涵盖过渡金属（钒、铁、铬）、硫、锌和氢的无机电对可提供定制化的电池电压，具体取决于正极液与负极液之间的热力学电位差。 (4:44)

Reading summary

Redox flow batteries present an alternative architecture for stationary grid storage by physically decoupling energy capacity from power output, enabling cost-effective duration scaling. According to Stanford Online, commercial adoption hinges on overcoming molecular solubility limits, active species crossover, and high membrane costs. Decoupling Power Output and Energy Capacity (0:00) Unlike conventional solid-state batteries that seal active materials within closed electrode cells, flow batteries dissolve redox-active molecules into circulating liquid electrolytes known as the anolyte and catholyte. This architecture separates power capability from energy capacity. Power delivery is governed strictly by the surface area and design of the central electrochemical cell stack, while total stored energy is dictated by the volume of external storage tanks and the concentration of the dissolved molecules. Because scaling storage duration requires only larger, relatively inexpensive storage tanks rather than additional reactor stacks, the technology provides a pathway to lower capital costs for stationary grids. This design trades volumetric footprint for cost efficiency, avoiding the tight gravimetric and packaging limits imposed on electric vehicles. Solubility Limits and Volumetric Energy Density (2:04) Because active charge carriers reside in liquid solutions, chemical solubility is the primary bottleneck for system energy density. A baseline aqueous flow battery operates at roughly 1 molar (1 M) concentration, with research targeting 2 M to 4 M. By comparison, pure water contains roughly 56 M of water molecules, charged lithium in graphite intercalation anodes reaches an equivalent concentration of 25 M, and pure metallic lithium hits 76 M. Due to this dilute molecular packing, flow batteries require substantially larger physical footprints than solid-state systems. Consequently, incremental synthetic improvements that increase solubility from 1 M to 2 M or 4 M yield significant practical dividends by reducing required tank volumes and balance-of-plant costs. Cycling Degradation and Selective Membrane Costs (3:27) Long-duration stationary storage requires molecules capable of withstanding thousands of charge and discharge cycles over decades. Any chemical breakdown, dimerization, or secondary irreversible reactions permanently deplete active material and cause capacity fade. Maintaining stability across continuous oxidation-state transitions is therefore critical. Simultaneously, the separator membrane between the anolyte and catholyte presents a major technical and economic challenge. It must act as a strict barrier to prevent large redox-active molecules from crossing over and causing self-discharge or cross-contamination, while maintaining high permeability for small counter-ions to preserve charge neutrality. Currently available commercial ion-selective separators, such as perfluorosulfonic acid membranes, remain among the most expensive components in the reactor stack. Inorganic Redox Chemistries and Voltage Pairing (4:44) System voltage, efficiency, and power output depend on pairing anolyte and catholyte chemistries with a wide thermodynamic spread between their standard reduction potentials. Inorganic options offer diverse mechanisms, including gaseous water splitting, zinc plating and stripping, sulfur-polysulfide conversion, and iron couples such as Fe²⁺/Fe³⁺ or Fe²⁺/Fe⁰. Transition metals like vanadium offer a particular operational advantage: vanadium can utilize multiple distinct oxidation states across both half-cells. This configuration helps mitigate permanent capacity loss, as cross-mixed active species can be chemically rebalanced rather than permanently contaminating the opposing electrolyte. While decoupling energy and power offers an advantageous model for multi-hour grid storage, the commercial viability of redox flow batteries remains constrained by chemical solubility limits, crossover-induced degradation, and the high manufacturing cost of ion-selective membranes.

Source quotes

[0:00] Holding energy in liquids rather than solid cells could change how storage scales. [0:00] Yi Cui: While conventional batteries lock their active materials inside sealed, solid-state electrode cells, flow batteries dissolve redox-active molecules directly into liquid electrolytes to store chemical energy. [0:00] Yi Cui: The total stored energy is dictated entirely by external factors: tank volume, solution solubility, and "the total number of redox molecules." [0:00] Yi Cui: By decoupling energy from power, scaling up duration does not require buying more expensive reactor stacks—operators simply install larger, inexpensive fluid storage tanks, giving the technology "just incredible value" and the potential for exceptionally low capital costs. [2:04] Yi Cui: Because liquid tanks carry the active materials, the chemical solubility of the redox species in the solvent serves as the primary bottleneck for energy density. [2:04] Yi Cui: Consequently, incremental synthetic improvements that increase flow battery solubility—elevating concentrations from 1 M to 2 M, or upwards to 4 M—deliver major dividends in shrinking system footprint and reducing tank costs. [3:27] Yi Cui: Because grid batteries are expected to operate for thousands of cycles over decades, the redox molecules must endure continuous, alternating oxidation and reduction without degrading. [3:27] Yi Cui: It has to act as an impermeable wall to large redox-active molecules while maintaining high ionic conductivity to allow small charge-balancing counter-ions to diffuse freely and preserve electrical neutrality. [3:27] Yi Cui: Currently available commercial ion-selective membranes (such as perfluorosulfonic acid membranes) are notoriously expensive, representing a significant percentage of the total stack cost and serving as a primary target for cost-reduction research. [4:44] Yi Cui: To construct a functioning cell, researchers pair a chemistry possessing a high standard reduction potential (serving as the positive catholyte) with a chemistry exhibiting a low reduction potential (serving as the negative anolyte).

Original transcript

[0:00] Redox flow is really important for  stationary energy storage. So the [0:06] basic idea of redox flow batteries  is to use soluble redox molecules to [0:15] carry the energy for both the anode and  cathode. Because they are all liquid, [0:21] they all flow in and out. That's why  we call these anolyte and catholyte. [0:29] It's both electrolyte and redox molecules  mixed together to form these liquid phases. [0:36] The major advantage is that the total amount  of energy is determined by the tank size to [0:44] hold the liquid. The total amount of energy  is the tank size, the solubility — basically, [0:51] the total number of redox molecules  determines the energy. And the power [0:58] depends on this electrode stack. When  you flow in the liquid during charging, [1:04] and during discharging you can flow the opposite  way, it's the stack size that determines the [1:12] power. You basically decouple the energy from the  power. So this potentially can be really low cost. [1:29] This idea has just incredible value  for stationary energy storage, [1:35] when your constraint is not the same as the mobile  power application. For mobile power you need high [1:43] energy density. The weight needs to be low and  the size needs to be small. But for stationary [1:51] energy storage, the system size could be  bigger. The only consideration is just to [1:57] make sure you can still meet the cost, given that  real estate cost needs to be taken into account. [2:04] So with redox flow batteries, the consideration  becomes solubility. You want it to be reasonably [2:14] high, otherwise your energy density will be  too low. Let's look at some numbers. If you [2:20] have one molar solubility, that means one mole  per liter. How does that compare with lithium [2:27] ion? How does it compare with other systems?  To calibrate yourself, one liter of pure water [2:37] is 56 moles of water molecules. If you look  at solid graphite in lithium ion batteries, [2:50] it's lithium carbon 6. We know the  density, so we can back calculate: [2:55] the molar concentration of lithium in graphite is  actually 25 molar. That's very high. Pure lithium [3:03] metal as an anode, in one liter, based on the  density you can calculate it's actually 76 molar. [3:10] That's very high. That's telling you solubility in  redox flow is important. One molar versus 25 and [3:19] 76 is very small. If you can increase from 1 molar  to 2 molar to 4 molar, that's very meaningful. [3:27] So that's the first consideration: solubility.  The second consideration is the stability of [3:33] redox molecules in solvent, because you  are going to do oxidation and reduction [3:38] back and forth so many cycles. You  want this molecule to be stable. If [3:44] electrochemically it is not stable, then  you're going to lose your capacity fast. [3:51] The third consideration is between  anolyte and catholyte, and separating [3:56] them with this membrane. During oxidation  and reduction, for most redox chemistry, [4:04] you do not want your redox molecules to be mixed.  If your catholyte and anolyte redox molecules mix, [4:11] that's not good. And you want a certain ion that  can diffuse through these membranes. So often you [4:19] need an ion-selective membrane to prevent other  species crossing over. You only want the ion to [4:27] cross over to balance the charge. Ion-selective  membranes are oftentimes high cost — the version [4:35] we now have available is high cost. So these are  the areas that require quite a bit of improvement. [4:44] Let's look at what redox reactions and molecules  you can use. This table gives you well-known [4:53] examples of inorganic redox reactions. You can see  water generating hydrogen could even be used — the [5:01] hydrogen reaction as a redox involving gas.  Zinc ions is one example. Sulfur ion. Fe or [5:11] Fe2 becomes metallic iron. Chromium, vanadium,  and so on. So this table is really summarizing [5:19] all these redox species, the inorganic ones.  They have different potentials, and you [5:25] need to pair a high potential one with a low  potential one to form the full battery cell.