Emerson Electric · Phase 1 Experimental Protocol

Aqueous Mg-Ion Battery
Laboratory Validation

A twelve-month protocol for reproducing, characterizing, and stress-testing the CuFe-PBA / COP aqueous battery chemistry at grid-relevant rates. Every measurement maps to an Emerson instrument.

12
months
4
go/no-go gates
2
parallel tracks
12
instruments
Based on: Chen et al., Nature Communications, February 2026
“Ultra-long cycle life aqueous batteries using MgCl₂/CaCl₂ water-in-salt electrolytes”
CuFe Prussian Blue Analogue Cathode · Covalent Organic Polymer Anode · 2.2V Full Cell
Confidential Shakopee, MN · Shell Space Phase 1
Contents

Protocol Sections

01

Scope & Objectives

This protocol governs the Phase 1 laboratory validation of an aqueous magnesium-ion battery based on the CuFe-PBA cathode / COP anode chemistry published by Chen et al. (Nature Communications, February 2026). The paper demonstrates 120,000+ cycles at 20 A/g with a 2.2 V full-cell voltage using a non-flammable, near-neutral pH electrolyte. Phase 1 answers the question the paper leaves open: does this chemistry survive at grid-relevant C-rates?

Primary Objectives

Track A — Chemistry Validation

Reproduce the full-cell architecture using pre-purified battery-grade MgCl₂ (Nedmag, Netherlands). Characterize electrochemical performance at C/10, C/4, C/2, and 1C. Establish degradation curves, copper leach rates, and impedance evolution over 1,000+ cycles at each rate.

Track B — Purification Development

Develop a repeatable, instrumented process to purify industrial-grade MgCl₂ (Intrepid Potash, Carlsbad NM) to battery-grade (≥99.9%) using Emerson DeltaV and Rosemount sensors. Validate via ICP analysis against Nedmag baseline. Qualify domestic electrolyte for cell production.

Success Criteria

ParameterTargetKill ThresholdInstrument
Capacity retention at C/4, 1,000 cycles >95% <80% NI HPS-17000
Coulombic efficiency >99.5% <95% NI HPS-17000
Electrolyte pH stability 4.91–7.02 <4.0 or >8.0 Rosemount 3900
Copper leach rate <5 ppm per 100 cycles >50 ppm per 100 cycles ICP-OES
Track B purity match ≥99.9% <99.5% ICP-OES
Impedance growth (EIS) <20% at 1,000 cycles >50% NI PXI
02

Safety Protocol & Hazard Assessment

The aqueous chemistry is intrinsically safer than lithium-ion (no flammable solvents, no thermal runaway). However, synthesis involves reagents that require standard chemical laboratory safety.

Hazard Inventory

HCN Risk — K₃[Fe(CN)₆]
Corrosive — NaOH (2M, Track B precipitation)
Irritant — CuSO₄, NMP
Inert Atmosphere — Argon Glovebox
Critical: Potassium Ferricyanide Handling

K₃[Fe(CN)₆] is stable and non-toxic under normal conditions. It must never contact strong acid (pH < 3). Acidification liberates hydrogen cyanide gas (HCN), which is lethal at >300 ppm. All ferricyanide work must occur in a fume hood with continuous airflow ≥100 fpm face velocity. Dedicated acid-free zone. Spill kit with sodium bicarbonate (NaHCO₃) immediately adjacent.

Layer Architecture

1
Facility
Fume hoods
Emergency shower
Eyewash station
Spill containment
2
Monitoring
Rosemount 928
H₂ gas detection
Rosemount 628-H₂
Continuous logging
3
Control
DeltaV SIS
Auto-shutdown
Interlock valves
Fisher regulators
4
Personnel
Lab coat, goggles
Nitrile gloves
Buddy system
WHMIS training
5
Emergency
Fire extinguisher
NaHCO₃ spill kit
Emergency phone
MSDS binder

PPE Requirements by Zone

ZoneMinimum PPEAdditional
Zone A — ElectrolyteLab coat, safety glasses, nitrile glovesFace shield when handling NaOH ≥1M
Zone B — SynthesisLab coat, splash goggles, nitrile glovesN95 respirator when handling K₃[Fe(CN)₆] powder; fume hood mandatory
Zone C — TestingLab coat, safety glassesHearing protection near environmental chamber compressor

Gas Monitoring Alarm Setpoints

GasSensorLow AlarmHigh AlarmAction
H₂Rosemount 928 / 628-H₂1,000 ppm2,500 ppmLOW: investigate, increase ventilation. HIGH: evacuate Zone B, DeltaV SIS shuts down electrolysis.
HCNDedicated electrochemical2 ppm5 ppmLOW: stop ferricyanide handling, verify acid segregation. HIGH: evacuate lab, initiate emergency protocol.
O₂Glovebox oxygen analyzer>1 ppm>5 ppmGlovebox breach. Purge. Do not open antechamber until <0.5 ppm.
03

Equipment Manifest

The core measurement stack is Emerson hardware. Two supplementary instruments (EQCM and vapor pressure osmometer) are sourced from specialty vendors; all primary process control and electrochemical characterization runs on Emerson-manufactured equipment.

Zone A — Electrolyte Preparation

Rosemount 228
Toroidal conductivity sensor. Monitors MgCl₂ concentration in real time during dissolution and purification. Range: 0–2,000,000 µS/cm. Non-contact measurement prevents electrode fouling in concentrated brine.
Continuous
Rosemount 3900
pH/ORP sensor with digital diagnostics. Validates electrolyte neutrality (pH 6.5–7.5) after preparation and tracks drift during cycling. Self-diagnostics flag sensor degradation before it affects readings.
Continuous
Rosemount 372
Dissolved oxygen sensor. Monitors O₂ levels during electrolyte degassing. Target: <0.5 ppm dissolved O₂ before cell fill. Excessive oxygen accelerates copper oxidation from the CuFe-PBA cathode.
QA Check
Micro Motion ELITE
Coriolis flow/density meter. Verifies electrolyte density corresponds to target MgCl₂ molarity (4.0M = ~1.295 g/mL, saturated 5.8M = ~1.36 g/mL at 25°C). Inline QA during batch preparation; primary QA for saturated batches where conductivity is non-monotonic. ±0.05% density accuracy.
Batch QA
Fisher Valves
Automated flow control on the DeltaV purification skid. Regulates NaOH addition rate, wash water flow, and recrystallization heating ramp. DeltaV-integrated for closed-loop process control.
Track B
DeltaV™ M-series
Distributed control system. Orchestrates the 5-stage purification skid (dissolution → precipitation → filtration → recrystallization → QA). Recipe-based batch control. Full historian logging. SIS-integrated safety interlocks.
Track B

Zone B — Cell Assembly & Chemistry

Argon Glovebox
Controlled atmosphere enclosure. O₂ < 1 ppm, H₂O < 1 ppm. Required for electrode handling and coin cell crimping. Eliminates moisture-induced side reactions during assembly.
Critical
Rosemount 928
Fixed gas detector for hydrogen. Monitors Zone B ambient during cell charging. H₂ evolution indicates water splitting (electrolyte stability window exceeded). DeltaV SIS input for automated shutdown.
Safety
Rosemount 628-H₂
Portable hydrogen detector. Complements the fixed 928 for spot checks near Swagelok cells and environmental chamber exhaust. 0–10,000 ppm range with T90 < 30 s response.
Safety

Zone C — Testing & Characterization

NI HPS-17000
Battery cycler, up to 150 kW. CC-CV, CP, CR, fast pulse protocols. The workhorse: runs 24/7 cycling at C/10, C/4, C/2, 1C. Multi-channel for parallel cell testing. EECOMOBILITY-integrated for AI-assisted degradation analysis.
24/7
NI PXI
Electrochemical impedance spectroscopy (EIS) platform. mHz to MHz frequency domain. Measures charge transfer resistance (Rct), Warburg diffusion, and electrode/electrolyte interface evolution. Baseline at formation, then every 100 cycles. Aqueous systems do not form a traditional SEI; EIS instead tracks interfacial passivation layers and copper deposition on electrode surfaces.
Periodic
EECOMOBILITY
AI-driven battery test software (NI/Emerson Ventures). Automated protocol management, real-time anomaly detection, degradation curve fitting, comparative analysis across C-rates. Handles automated test scheduling so the electrochemists focus on chemistry, not data plumbing.
Software
ICP-OES
Inductively coupled plasma optical emission spectrometer. Quantifies dissolved Cu, Fe, Mg, Ca, K in cycled electrolyte at sub-ppb detection limits. The primary instrument for tracking cathode dissolution kinetics.
Periodic
Environmental Chamber
Temperature-controlled cycling enclosure. −20°C to +60°C. Required for thermal qualification: nominal cycling at 25°C, then stress tests at 0°C, 40°C, and 55°C. Validates electrolyte stability window across deployment climates.
Thermal
EQCM
Electrochemical quartz crystal microbalance. Measures nanogram-scale mass changes at the electrode surface during cycling in real time. Quantifies copper deposition rate on the anode and active material dissolution from the cathode — the in-situ complement to ICP’s periodic sampling. Coupled with the NI PXI potentiostat.
Advanced
Vapor Pressure Osmometer
Measures water activity (aw) of the concentrated electrolyte. Confirms the water-in-salt regime (target aw < 0.55). The single measurement that validates the 2.2 V electrochemical stability window.
QA
04

Reagents & Materials

All reagents are commodity chemicals. No controlled substances, no conflict minerals, no export-restricted materials. Total Phase 1 materials cost: <$28,000.

Electrolyte

Magnesium Chloride Hexahydrate
MgCl₂·6H₂O
Track A: Nedmag (Veendam, NL), ≥99.9% battery-grade
Track B: Intrepid Potash (Carlsbad, NM), ~95–97% industrial
Target: 4.0M and saturated (~5.8M) solutions in deionized water (parallel batches per Concentration Matrix)
Qty: 75 kg Track A + 250 kg Track B (uplift covers parallel saturated batches)
Storage: Sealed containers, desiccated, RT
Deionized Water
H₂O
Grade: Type I (18.2 MΩ·cm), ASTM D1193
TOC: <5 ppb
Source: In-house DI system (building utility)
Use: Electrolyte dissolution, washing, rinsing

Cathode Precursors

Copper Sulfate Pentahydrate
CuSO₄·5H₂O
Grade: ACS Reagent, ≥99.0%
Source: Sigma-Aldrich or Freeport-McMoRan (Phoenix, AZ)
Qty: 5 kg
Hazard: Irritant. Harmful if swallowed. Aquatic toxicant.
Storage: Sealed, dry, RT
Potassium Ferricyanide
K₃[Fe(CN)₆]
Grade: ACS Reagent, ≥99.0%
Source: Sigma-Aldrich
Qty: 5 kg
Hazard: Low toxicity. NEVER MIX WITH ACID.
Storage: Sealed, light-protected, RT. Separate from acids.

Anode Precursors

COP Monomer
Per Chen et al. synthesis
Type: Triazine/imine-linked organic precursor
Grade: Synthesis-grade, ≥98%
Source: Specialty chemical supplier (TCI, Alfa Aesar)
Qty: 2 kg
Note: Exact monomer identity per published supplementary information
Solvothermal Solvent
DMSO or DMF
Grade: Anhydrous, ≥99.9%
Source: Sigma-Aldrich
Qty: 10 L
Hazard: Skin permeable (DMSO), reproductive toxin (DMF)
Storage: Flammable cabinet, under nitrogen blanket

Electrode Components

Carbon Black (Super P)
C
Grade: Conductive, BET ≥60 m²/g
Source: Imerys (formerly Timcal)
Qty: 2 kg
Role: Conductive additive in electrode slurry (20 wt%)
Storage: Sealed, desiccated, RT
PVDF Binder
Polyvinylidene fluoride
Grade: Kynar HSV 900, MW ~1,000,000
Source: Arkema
Qty: 1 kg
Role: Electrode binder (10 wt%)
Solvent: NMP (N-methyl-2-pyrrolidone)
NMP
N-methyl-2-pyrrolidone
Grade: Anhydrous, ≥99.5%
Source: Sigma-Aldrich
Qty: 20 L
Hazard: Reproductive toxin. Fume hood mandatory.
Storage: Flammable cabinet, sealed, RT
Current Collectors
Ti foil (cathode) / SS foil (anode)
Cathode: Ti foil, 25 µm, ≥99.6% (aqueous-stable)
Anode: Stainless steel 316L, 25 µm
Source: MTI Corp or Goodfellow
Note: Al/Cu current collectors corrode in aqueous electrolyte
Separator
Glass fiber or hydrophilic PE
Type: Whatman GF/A glass fiber, 260 µm
Alternative: Celgard hydrophilic-treated PE
Source: Cytiva (GF/A) or Celgard (Charlotte, NC)
Qty: 500 sheets (coin cell) + 100 sheets (prismatic)
Coin Cell Hardware
CR2032 + Swagelok
Screening: CR2032 cases, spacers, springs, wave washers
Cycling: Swagelok-type 2-electrode cells (PTFE body)
Source: MTI Corp / Swagelok
Qty: 200 CR2032 kits + 20 Swagelok cells

Purification Reagents (Track B)

Sodium Hydroxide
NaOH
Grade: ACS Reagent, ≥97%
Concentration: 2M solution for selective precipitation
Qty: 25 kg
Hazard: Corrosive. Face shield + gloves required.
Purpose: Precipitate Ca²⁺, Fe³⁺, heavy metal impurities
Type 4A Zeolite
Na₂O·Al₂O₃·2SiO₂·nH₂O
Grade: Ion exchange, 8–12 mesh beads
Source: Sigma-Aldrich or UOP
Qty: 50 kg
Purpose: Selective Ca²⁺/Sr²⁺ removal from MgCl₂ brine
Regeneration: 10% NaCl brine wash, ≥50 cycles
The paper proved
120,000
cycles at 20 A/g
This protocol asks
What happens at C/4?
Grid storage operates at C/4 to C/2. Nobody has published this data.
The first lab to produce it owns the answer.
Reference Figures

Cell Architecture & Expected Signatures

Three reference diagrams for the Phase 1 team. The cell schematic defines the physical stack order. The voltage profile shows what a healthy charge/discharge curve looks like. The Nyquist plot shows what healthy impedance looks like and how degradation manifests.

CR2032 Cell Stack

BOTTOM CASE (cathode side) Ti FOIL (25 µm) CuFe-PBA CATHODE 70:20:10 AM:CB:PVDF · 50-80 µm ~80 µL electrolyte GF/A SEPARATOR Whatman · 260 µm · 16 mm ø ~80 µL electrolyte COP ANODE 70:20:10 AM:CB:PVDF · N/P 1.1-1.2 SS 316L (25 µm) SPACER (0.5 mm) WAVE SPRING TOP CAP + GASKET CR2032 · 20 mm ø × 3.2 mm 2.2 V CELL STACK ORDER

Voltage Profile — What Good Looks Like

2.2 1.8 1.4 1.0 0.6 0.2 Voltage (V) Capacity (mAh/g) 0 25 50 Charge Discharge CV hold Voltage hysteresis (should be <200 mV) EXPECTED VOLTAGE PROFILE (C/10 FORMATION)

Nyquist Plot — Healthy vs. Degraded

-Z'' (Imaginary / Ohm) Z' (Real / Ohm) Rₚ R₞ₜ (healthy) R₞ₜ (degraded — investigate) Warburg (diffusion) Cycle 0 (baseline) After 1,000 cycles (R₞ₜ growth >50% = investigate) EIS NYQUIST PLOT — 100 kHz TO 10 mHz

Schematic representations. Actual impedance magnitudes will be determined during formation (Procedure F, Step 03).

05

Procedure A: Electrolyte Preparation

Two parallel tracks (A: Nedmag-purified, B: domestic-purified) × two parallel concentrations (4.0M handling-optimized + saturated 5.8M paper-replication). Chen et al. characterizes the cell at saturated MgCl₂ (~5.8M at 25°C), with 1M and 3M tested for water-in-salt regime comparison. The proposal’s 4.0M handling-optimized concentration sits below precipitation-risk threshold and remains in the WIS regime (aw < 0.55). Phase 1 tests both side-by-side as a publishable Phase 1 deliverable: establish the optimal grid-scale concentration for industrial deployment. All target pH 6.5–7.5.

Concentration Matrix · Phase 1 Deliverable
Parameter4.0M (handling-optimized)Saturated ≈5.8M (paper-replication)
Mass MgCl₂·6H₂O per 1 L H₂O813.2 g~1180 g (saturation limit)
Density @ 25°C1.290–1.300 g/mL1.355–1.370 g/mL
Conductivity @ 25°C180,000–220,000 µS/cm~250,000–310,000 µS/cm (decreases above 4M due to viscosity)
Water activity (aw)0.45–0.550.32–0.40
Viscosity @ 25°C (cP)~9–12~28–40
Precipitation risk < 15°CLowModerate — warm to 25°C before use
Cells in matrix~46 cells (Tracks A+B, all C-rates)~46 cells (Tracks A+B, all C-rates, paper-direct comparison)

Why test both: (1) saturated reproduces Chen exactly — the calibration cell that says “our hands match the paper’s hands.” (2) 4.0M is the engineering candidate — lower viscosity means better ion transport at scale, easier industrial handling, and lower precipitation risk in cold climates. Phase 1 deliverable: a head-to-head cycle-life and round-trip-efficiency comparison at C/4 and C/2. The optimal grid-scale concentration becomes the answer the rest of the field will cite.

Electrolyte Source Strategy

What Chen et al. used. Saturated MgCl₂ (~5.8M at 25°C) is the headline electrolyte in the published paper, with 1M and 3M tested for comparison (SI Figs. 13–14). The COP polymer family is Hex-TAPZ, Hex-DABZ, and Hex-TADD; Hex-TADD-COP is the headliner that delivers the 120,000-cycle result. Source supplier and reagent grade are not disclosed in the publicly accessible Supplementary Information; standard Chinese university convention is Aladdin, Macklin, or InnoChem analytical-reagent grade (≥99% MgCl₂·6H₂O hexahydrate). Phase 1 verifies this against the supplier’s certificate of analysis before using.

Track A — Nedmag (Netherlands), the safe baseline. Solution-mined from the Veendam deposit in eastern Netherlands. Reserves: estimated 1.5–2.0 billion tonnes MgCl₂-equivalent. Current production: ~440,000 tonnes/yr salt across the company’s product mix. At full Phase 3 scale (3–5% global stationary storage market capture), Emerson demand projects ~50,000–150,000 tonnes/yr electrolyte-grade salt. Nedmag reserves support >100 years at that rate. “Running out” is not a Phase 1–4 concern.

Track B — Intrepid Potash (Carlsbad, NM), the US security play. Industrial-grade MgCl₂ (90–98% purity) from existing US potash operations. Phase 1 Track B establishes the DeltaV purification protocol that lifts this to ≥99.9% (Nedmag-equivalent). Closes the geopolitical gap; eliminates Netherlands single-source dependency for US deployments.

Track C — Desalination brine, the inexhaustible source. Global desalination plants discharge ~142 million m³/day of MgCl₂/CaCl₂-rich brine, almost all dumped to ocean. Capturing 0.1% of this stream supplies Phase 4 demand perpetually. Track C economics depend on Phase 1 Track B purification protocol working at scale — which is why Track B is committed in Phase 1, not deferred.

Summary: Nedmag alone covers Phase 1–4 with multi-decade headroom. Track B closes the geopolitical gap. Track C makes the supply effectively infinite. Salt is not the constraint — getting it pure enough is the constraint, and that is exactly the engineering Emerson is built to do.

Track A — Battery-Grade Electrolyte

Target Specification

4.0M
Concentration
6.5–7.5
pH
1.295
Density (g/mL)
01
Dissolution
Two parallel batches.

4.0M batch: Weigh 813.2 g MgCl₂·6H₂O (Nedmag, ≥99.9%) into a 2 L borosilicate beaker. Add 1000 mL Type I deionized water. Stir at 400 rpm with PTFE-coated magnetic stir bar until fully dissolved (~15 min at RT).

Saturated (~5.8M) batch: Weigh ~1180 g MgCl₂·6H₂O into a separate 2 L beaker. Add 1000 mL DI water. Stir at 400 rpm warmed to 30–35°C on a hotplate until fully dissolved (~30–45 min). Cool to 25°C before transferring; if any precipitate appears on cooling, gently warm and re-stir (this is the saturation boundary — expected at the 5.8M target).

Both dissolutions are endothermic; solution temperature will drop 5–8°C unless heated. Label batches A-4M-NNN and A-SAT-NNN.

MW MgCl₂·6H₂O = 203.30 g/mol. 4.0M in 1 L = 4.0 × 203.30 = 813.2 g. Saturated at 25°C ≈ 5.8M = 5.8 × 203.30 = 1179.1 g (use ~1180 g target).
02
QA — Conductivity
Immerse Rosemount 228 toroidal conductivity probe. Record reading at 25°C. Expected: 180,000–220,000 µS/cm for the 4.0M batch; 250,000–310,000 µS/cm for the saturated batch (conductivity peaks near 4M and decreases above due to viscosity dominating ion mobility). If outside range, verify mass/volume and re-dissolve. Log to DeltaV historian.
03
QA — pH
Measure with Rosemount 3900 pH sensor (2-point calibrated with pH 4.00 and 7.00 buffers within 24 h). Target: pH 6.5–7.5. If pH < 6.5, add 0.1M NaOH dropwise with stirring until target reached. If pH > 7.5, add 0.01M HCl dropwise. Record final pH.
04
QA — Density
Measure inline with Micro Motion ELITE Coriolis meter or benchtop densitometer. Target: 1.290–1.300 g/mL for the 4.0M batch and 1.355–1.370 g/mL for the saturated batch, both at 25°C. Density confirms molarity independent of conductivity measurement (and is the primary QA on saturated batches because conductivity is non-monotonic). If out of spec, adjust water volume and re-verify.
05
QA — Water Activity
Measure water activity (aw) with a vapor pressure osmometer or chilled-mirror hygrometer (e.g., METER AquaLab). Target: aw = 0.45–0.55 at 25°C. This confirms the “water-in-salt” regime: at aw < 0.55, free water molecules are coordinated to Mg²⁺ ions, suppressing the hydrogen evolution reaction and expanding the electrochemical stability window past 2.2 V. If aw > 0.7, the electrolyte is too dilute and water will split at ~1.23 V under applied potential.

This is the measurement that validates the entire electrochemical premise. Without it, the 2.2 V voltage window is an assumption, not a fact.
06
Degassing
Transfer electrolyte to Schlenk flask. Apply vacuum (50–100 mbar) under gentle stirring for 30 min at RT. Alternatively, sparge with ultrapure argon (99.999%) at 100 mL/min for 45 min. Verify dissolved O₂ < 0.5 ppm with Rosemount 372 before sealing under Ar.
07
Baseline ICP
Withdraw 10 mL aliquot. Submit for ICP-OES analysis: Mg, Ca, Cu, Fe, K, Na, Sr, Ba, Mn, Zn. This is the purity baseline. All postcycling ICP results will be compared to this reference. Record as LOT-A-XXXX.

Track A electrolyte should show <1 ppm total transition metals. Any reading >10 ppm indicates contaminated reagent — reject lot.
08
Storage
Store in amber borosilicate bottles under argon headspace. Cap tightly with PTFE-lined caps. Label with lot number, date, molarity, pH, density. Shelf life: 6 months sealed; re-verify conductivity and pH before use if >30 days old.
06

Procedure B: CuFe-PBA Cathode Synthesis

Coprecipitation of copper(II) sulfate and potassium ferricyanide to form the open-framework Prussian Blue Analogue Cu₂[Fe(CN)₆]. Room-temperature synthesis. No specialized atmosphere required (aqueous reaction).

Reaction

3 CuSO₄ + 2 K₃[Fe(CN)₆] → Cu₃[Fe(CN)₆]₂ + 3 K₂SO₄
(simplified; actual stoichiometry depends on Cu:Fe ratio and crystallization conditions)
01
Prepare Solution A — Copper Sulfate
Dissolve 24.97 g CuSO₄·5H₂O in 1000 mL deionized water to produce a 0.1M solution. Stir until fully dissolved (~10 min). Solution should be clear blue with no particulates.

MW CuSO₄·5H₂O = 249.69 g/mol. For 0.1M in 1 L: 0.1 × 249.69 = 24.97 g.
02
Prepare Solution B — Potassium Ferricyanide
Dissolve 22.06 g K₃[Fe(CN)₆] in 1000 mL deionized water to produce a 0.067M solution. Stir until fully dissolved. Solution should be clear deep yellow/orange. Fume hood mandatory. No acids in vicinity.

MW K₃[Fe(CN)₆] = 329.24 g/mol. For 0.067M in 1 L: 0.067 × 329.24 = 22.06 g.
03
Coprecipitation
Using a peristaltic pump (or burette), add Solution B dropwise into Solution A at 2–5 mL/min under vigorous stirring (800–1000 rpm). A brown/reddish-brown precipitate forms immediately. Maintain temperature at 25 ± 2°C throughout. Total addition time: ~3.5–8 hours.

Slow addition rate controls particle size and crystallinity. Faster addition produces amorphous material with higher initial capacity but faster degradation. This is a tunable parameter for Phase 1 optimization.
04
Aging
After complete addition, continue stirring at 400 rpm for 12–24 hours at RT. This ripening step improves crystallinity and reduces lattice vacancies in the PBA framework. Longer aging generally improves cycle stability.
05
Washing & Filtration
Collect precipitate by vacuum filtration on 0.45 µm PVDF membrane. Wash with 3 × 500 mL deionized water (removes K₂SO₄ byproduct). Wash with 1 × 200 mL ethanol (removes residual water). Verify wash water conductivity < 50 µS/cm with Rosemount 228 before stopping.
06
Drying
Transfer filter cake to vacuum oven. Dry at 80°C, <100 mbar for 12 hours. Do not exceed 120°C — PBA framework decomposes above ~200°C but begins losing coordinated water at lower temperatures. Final product: fine brown powder. Expected yield: ~8–12 g per batch.
07
Characterization
Submit 200 mg for:
XRD (powder X-ray diffraction) — confirm face-centered cubic PBA structure (Fm3̄m), lattice parameter ~10.1 Å
TGA (thermogravimetric analysis) — verify coordinated water content (expected ~25–35 wt% loss below 200°C)
ICP-OES — confirm Cu:Fe ratio (~3:2 for Cu₃[Fe(CN)₆]₂)
BET — surface area (expected 20–60 m²/g depending on synthesis conditions)

Label as CuFe-PBA-XXXX (batch number). Log all characterization data.
07

Procedure C: COP Anode Synthesis

Solvothermal condensation of the covalent organic polymer. The COP stores charge through reversible radical cation formation — no intercalation, no volume change, no cracking. The absence of structural deformation during cycling is the mechanism behind the published 120,000-cycle result.

Polymer Family · Three Variants, One Headliner

Chen et al. characterizes three covalent organic polymer variants from a common hexaketone monomer (Hex):

Phase 1 prepares Hex-TADD-COP for the primary cell matrix (Track A). Hex-TAPZ and Hex-DABZ are synthesized in smaller batches as comparison sets for ranking degradation pathways under the same C-rate sweep. Naming the specific variant matters because peer reviewers will ask, and because process-IP downstream (electrode formulation, binder, current density windows) is polymer-specific.

Charge Storage Mechanism

Unlike intercalation anodes (graphite, Li₄Ti₅O₁₂), the Hex-COP family stores charge via reversible radical cation formation at C=N active sites along the polymer backbone. The backbone accepts and releases electrons without structural deformation. No expansion, no contraction — the degradation mechanism that limits every intercalation-based battery does not exist here.

01
Monomer Preparation
Weigh hexaketone (Hex) monomer plus the corresponding diamine: tetraaminodibenzo-p-dioxin (TADD) for primary cells, 1,4-diaminobenzene (DABZ) and 2,3,5,6-tetraaminopyrazine (TAPZ) for comparison cells. Stoichiometry per Chen et al. SI Fig. 1 (Schiff-base polycondensation). Typical batch size: 5–10 g total per polymer. Verify purity by NMR (solid-state 13C and 1H per SI Figs. 5–7) or melting point before use. Monomers must be dry (<500 ppm H₂O by Karl Fischer).
02
Solvothermal Condensation
Combine monomer with 50–100 mL anhydrous DMSO (or DMF per paper) in a PTFE-lined stainless steel autoclave (250 mL capacity). Seal under argon. Heat to 180°C at a ramp rate of 2°C/min. Hold at 180°C for 72 hours.

Temperature and time are critical. 180°C drives condensation without charring. Shorter reaction times produce oligomers with lower capacity. Longer times risk crosslinking beyond optimal porosity.
03
Cooling & Recovery
Cool autoclave to RT at natural rate (do not quench — thermal shock can crack the polymer network). Open in fume hood. Collect dark brown/black solid product by vacuum filtration on 0.45 µm membrane.
04
Purification
Soxhlet extraction with methanol for 24 hours, then THF for 24 hours. This removes unreacted monomer, oligomeric fragments, and trapped solvent. Alternatively: wash sequentially with DMF (3×), methanol (3×), acetone (2×), deionized water (2×).
05
Drying & Activation
Dry in vacuum oven at 120°C, <50 mbar, 24 hours. The higher temperature (vs. PBA) is safe here — COPs are thermally stable to ~350°C. This step also activates the pore structure by removing trapped solvent molecules.
06
Characterization
Submit 200 mg for:
FTIR — confirm C=N imine linkage (~1620 cm⁻¹), absence of aldehyde/amine peaks
Solid-state ¹³C NMR — verify triazine/imine connectivity
TGA — thermal stability onset, residual solvent content
BET — surface area (expected 200–800 m²/g, depending on framework)
Elemental analysis — C, H, N composition vs. theoretical

Expected yield: 3–7 g per batch. Label COP-XXXX.
08

Procedure D: Electrode Fabrication

Standard slurry casting on metallic current collectors. The composition ratio — 70:20:10 active material : carbon black : PVDF — follows the paper. Coating uniformity directly determines capacity uniformity and cycle life.

Emerson Relevance

Electrode slurry viscosity is a Rosemount measurement. Coating thickness uniformity at scale maps directly to Afag linear motion systems for doctor blade or slot-die application. This procedure is where lab science meets manufacturing instrumentation.

Cathode Electrode (CuFe-PBA)

01
PVDF Dissolution
Dissolve 0.10 g PVDF (Kynar HSV 900) in 2.0 mL NMP. Stir at 60°C for 2–4 hours until completely dissolved. Clear, viscous solution. Fume hood. NMP is a reproductive toxin.
02
Slurry Mixing
Add 0.70 g CuFe-PBA and 0.20 g Super P carbon black to the PVDF/NMP solution. Mix with planetary ball mill or high-shear mixer at 2000 rpm for 30 min. Result: homogeneous dark slurry with no visible agglomerates. If too thick, add NMP in 0.2 mL increments. Target viscosity: 2,000–5,000 cP.

Mass ratio: 0.70 : 0.20 : 0.10 = 70 : 20 : 10 (AM : CB : PVDF). Scale linearly for larger batches.
03
Casting
Cast slurry onto Ti foil (25 µm) using doctor blade set to 200 µm wet film thickness. Draw speed: 10–20 mm/s. Wet film should be uniform with no streaks, pinholes, or edge beading. For coin cells, cast a 100×100 mm area.

Titanium foil is required. Copper and aluminum corrode in aqueous MgCl₂ electrolyte.
04
Drying
Pre-dry on hotplate at 60°C for 2 hours (evaporates bulk NMP). Transfer to vacuum oven: 120°C, <100 mbar, 12 hours. Final electrode thickness (active layer): 50–80 µm. Active material loading target: 1.5–3.0 mg/cm².
05
Calendering & Punching
Calender (roll press) at 2–5 MPa to target porosity 35–45%. Punch electrodes: 12 mm diameter for CR2032, 16 mm diameter for Swagelok. Weigh each electrode (±0.01 mg). Calculate active material mass. Record in electrode log.

Anode Electrode (COP)

Identical procedure (steps 01–05) with the following substitutions:

ParameterCathode (CuFe-PBA)Anode (COP)
Active materialCuFe-PBA powderCOP powder
Current collectorTi foil, 25 µmSS 316L foil, 25 µm
Loading target1.5–3.0 mg/cm²1.0–2.5 mg/cm² (capacity-matched)
Doctor blade gap200 µm150–200 µm (adjust for capacity balance)

Capacity Balancing — Critical

The anode capacity must be 10–20% excess vs. cathode (N/P ratio 1.1–1.2). If the cathode delivers more charge than the anode can accept, metallic Mg deposits on the polymer surface rather than inserting into the C=N framework — irreversible capacity loss and potential dendrite nucleation. Calculate theoretical capacity from published specific capacity values, weigh both electrodes, and select matched pairs before assembly.

09

Procedure E: Cell Assembly

Two formats: CR2032 coin cells for rapid screening and Swagelok cells for long-term cycling. All assembly in the argon glovebox (O₂ < 1 ppm, H₂O < 1 ppm). Note: The electrolyte is aqueous, but the glovebox protects the COP anode from ambient oxygen, which irreversibly oxidizes the radical cation sites. Electrodes enter the glovebox dry; the degassed electrolyte is introduced at the last step via micropipette. The water is controlled, not excluded.

CR2032 Coin Cell (Screening)

01
Glovebox Entry
Transfer electrodes, separator (Whatman GF/A, 16 mm diameter), electrolyte (sealed vial), coin cell components (case, spacer, spring, gasket) into glovebox antechamber. Purge 3 cycles (vacuum to 0.1 mbar, backfill Ar to 800 mbar). Verify O₂ < 1 ppm on glovebox analyzer before opening inner door.
02
Stack Assembly
Bottom case (cathode side) → CuFe-PBA cathode (active side up) → 2 drops (~80 µL) electrolyte → GF/A separator → 2 drops electrolyte → COP anode (active side down) → SS spacer (0.5 mm) → wave spring → top cap with gasket.

Electrolyte volume is critical. Too little: dry spots, uneven current distribution. Too much: hydraulic pressure on seal, leakage. 160 µL total for CR2032 with GF/A separator is the starting point — optimize empirically.
03
Crimping
Crimp with hydraulic coin cell crimper at 750–1000 psi. Consistent pressure is essential for reproducibility. Inspect seal visually — no visible electrolyte leakage. Measure OCV immediately: expected 0.8–1.2 V for freshly assembled cell.
04
Rest & Wetting
Rest cell for 6–12 hours at OCV before any cycling. This allows electrolyte to fully wet the separator and electrode pores. Monitor OCV during rest — a stable or slowly rising OCV indicates proper wetting. A dropping OCV suggests internal short or electrolyte decomposition.

Swagelok Cell (Long-Term Cycling)

Identical electrode/separator/electrolyte stack in a PTFE-body Swagelok union fitting. Benefits: easy disassembly for postmortem analysis, larger electrolyte volume (~1–2 mL), better seal for months-long cycling. Use 16 mm diameter electrodes with 18 mm GF/A separator.

Cell Naming Convention

Format: [TYPE]-[TRACK]-[BATCH]-[SEQ]
Example: SW-A-0301-04 = Swagelok, Track A electrolyte, March Batch 01, Cell #4
Example: CR-B-0602-11 = CR2032, Track B electrolyte, June Batch 02, Cell #11

Cell Build Matrix — Phase 1

PurposeFormatQtyC-RateTrack
Formation screeningCR203220C/10A
C-rate sweepSwagelok12C/10, C/4, C/2, 1C (3 each)A
Long-term cyclingSwagelok8C/4 (4), C/2 (4)A
Temperature studySwagelok8C/4 @ 0/25/40/55°C (2 each)A
Track B qualificationSwagelok8C/4 (4), C/2 (4)B
Electrolyte reuse testSwagelok4C/4A→reuse
High-precision coulometrySwagelok4C/4 (2), C/2 (2)A
Self-discharge / calendarSwagelok4C/4 charge, 72h restA
EQCM (in-situ mass)EQCM cell4C/10 (slow, mass-resolved)A
Postmortem reserveCR203216VariousA
Total92
10

Procedure F: Formation & Conditioning

Slow initial cycling establishes a stable electrode/electrolyte interface. In aqueous systems there is no traditional SEI layer, but the first cycles activate the PBA framework and establish the COP radical cation equilibrium.

01
Formation Cycling (C/10)
Cycle at C/10 between 0.0 V and 2.2 V for 5 complete cycles on the NI HPS-17000. Mode: CC-CV (constant current → constant voltage at upper cutoff, hold until current drops to C/50). Record capacity, coulombic efficiency, and voltage profile for each cycle.

C/10 = capacity/10 hours. For a ~50 mAh/g material at 2 mg loading: ~10 µA. The HPS-17000 handles this at its low-current module.
02
Formation QA
After 5 formation cycles, evaluate:
• Coulombic efficiency cycles 2–5: must be >98% (cycle 1 is typically 70–90% due to irreversible reactions)
• Capacity stabilization: cycles 3–5 should vary <5%
• No anomalous voltage plateaus (would indicate side reactions or short circuit)

KILL: If CE <90% on cycle 5, or capacity drops >20% between cycles 3–5, flag cell for postmortem. Do not continue to rate testing.
03
Baseline EIS
After formation, perform EIS on the NI PXI platform:
• Frequency range: 100 kHz to 10 mHz
• AC amplitude: 10 mV (potentiostatic mode)
• At 50% SOC (discharge to mid-voltage, rest 1 hour)

Record Nyquist plot. Extract: Rs (series resistance), Rct (charge transfer resistance), W (Warburg coefficient). This is the cycle 0 impedance baseline. All future EIS is compared to this reference.
04
Baseline ICP (Electrolyte)
For Swagelok cells: withdraw 50 µL electrolyte sample via syringe port. Submit for ICP-OES. Compare to pre-cycling baseline (Procedure A, Step 06). Record Cu, Fe, K concentrations. This establishes the initial copper leach rate from formation cycling.
11

Procedure G: C-Rate Testing Matrix

The central experiment. Four C-rates, each run to 1,000+ cycles. No group has published rate-dependent cycling data for this chemistry. These measurements determine whether the lab result translates to a grid product.

C/10
BASELINE
10h charge/discharge
C/4
GRID MINIMUM
4h charge/discharge
C/2
GRID STANDARD
2h charge/discharge
1C
PEAK SHAVING
1h charge/discharge

Protocol

01
Cycling Parameters
All cycling on NI HPS-17000 with EECOMOBILITY data acquisition:
• Mode: CC-CV charge / CC discharge
• Voltage window: 0.0 V – 2.2 V
• CV hold at 2.2 V until current drops to C/20
• Rest period between charge/discharge: 5 min
• Temperature: 25 ± 1°C (environmental chamber)
• Data logging: voltage, current, capacity at 1 Hz (every datapoint)
02
Rate Sweep (First 50 Cycles)
For each cell: 5 cycles at each rate in ascending order:
5 @ C/105 @ C/45 @ C/25 @ 1C5 @ 2C5 @ C/10 (recovery check)

This produces the rate capability curve. The recovery C/10 must return to >95% of initial C/10 capacity. If not, permanent damage occurred at higher rates.
03
Sustained Cycling (Cycles 50–1,000+)
Assign each cell group to its target rate and cycle continuously:
• Group 1 (3 cells): C/10 — baseline, ~500 cycles in 12 months
• Group 2 (3 cells): C/4 — grid minimum, ~1,500 cycles in 12 months
• Group 3 (3 cells): C/2 — grid standard, ~3,000 cycles in 12 months
• Group 4 (3 cells): 1C — peak shaving, ~6,000 cycles in 12 months
04
Monitoring Cadence
EECOMOBILITY runs automated anomaly detection on every cycle. Manual review weekly. Specific data pulls:
Every 100 cycles: EIS (NI PXI), dQ/dV analysis, capacity check at C/10
Every 500 cycles: ICP electrolyte sample, pH check, full capacity/efficiency report
Every 1,000 cycles: Comprehensive review with degradation curve fitting, extrapolation to EOL

Cycle Life Projection Methodology

After 1,000 cycles at each rate, fit capacity vs. cycle number to both linear and power-law degradation models. Extrapolate to 80% capacity retention (industry-standard EOL for BESS). Compare extrapolated cycle life at C/4 against the 120,000 cycles demonstrated at 20 A/g. The ratio defines the practical-vs-laboratory performance discount factor — this single number determines commercial viability.

12

Procedure H: Characterization Schedule

Every characterization technique below has a defined cadence, a target parameter, and an action threshold. The schedule is designed to catch degradation mechanisms before they become ambiguous.

TechniqueInstrumentCadenceWhat It Reveals
EIS NI PXI Every 100 cycles Charge transfer resistance evolution. Early warning for electrode degradation, electrolyte decomposition, or interface passivation. Rct growth >50% triggers investigation.
ICP-OES ICP-OES Every 500 cycles Dissolved Cu, Fe, K, Mg, Ca in electrolyte. Quantifies cathode dissolution rate. Cu accumulation trajectory predicts remediation interval.
pH Rosemount 3900 Every 100 cycles Electrolyte acidification/basification trend. Drift outside 4.91–7.02 range indicates uncontrolled side reactions.
Conductivity Rosemount 228 Every 500 cycles Electrolyte concentration stability. Significant drift indicates water loss (evaporation from imperfect seal) or salt consumption.
dQ/dV Analysis NI HPS-17000 Every 100 cycles Differential capacity plots reveal changes in redox peak positions and intensities. Phase changes, new side reactions, and loss of active material all have distinct dQ/dV signatures.
XRD (postmortem) External lab M6, M12 Crystal structure of cycled CuFe-PBA cathode. Confirms framework integrity, detects amorphization or phase transitions. Compared to as-synthesized baseline.
SEM/EDX (postmortem) External lab M6, M12 Electrode morphology, particle cracking, copper deposition patterns, binder distribution. EDX maps element distribution across the electrode cross-section.
Rate Capability Check NI HPS-17000 Every 500 cycles 5-cycle sweep at C/10 to verify recoverable capacity hasn’t degraded independently of rate effects.
GITT NI HPS-17000 M1, M3, M6, M9, M12 Galvanostatic Intermittent Titration Technique. 10-min current pulse at C/10, 60-min OCV relaxation, repeat across full SOC range. Yields Mg²⁺ apparent diffusion coefficient (Dapp) in both electrodes as a function of SOC and cycle age. The rate-limiting step at C/4 vs. 20 A/g is almost certainly diffusion — GITT quantifies it directly.
High-Precision Coulometry NI HPS-17000 (HPC mode) Continuous (dedicated cells) Coulombic efficiency measured to ≥4 decimal places (e.g., 99.95% vs. 99.99%). At 120,000 cycles, even 0.01% parasitic loss per cycle compounds to 100% capacity loss at cycle ~7,000. Two dedicated cells at C/4 and C/2 run HPC continuously. This is the Dahn method — the only way to predict multi-decade lifetime from 12 months of data.
Self-Discharge / Calendar Aging NI HPS-17000 M3, M6, M9, M12 Charge to 100% SOC, rest at OCV for 72 hours (25°C), measure residual capacity by full discharge. Grid batteries idle at partial SOC for hours between dispatch cycles. Self-discharge >5%/month at 25°C is a red flag for parasitic side reactions consuming Mg²⁺ or water.
Water Activity (aw) Vapor pressure osmometer Every batch + M3, M6 Confirms the “water-in-salt” regime. At 4.0M MgCl₂, expected aw ≈ 0.45–0.55. If aw > 0.7, free water is available to split at 1.23 V and the expanded stability window claim does not hold. Measured on fresh and postcycling electrolyte to detect dilution from water ingress.
Raman Spectroscopy External lab (confocal Raman) M3, M6, M12 Ex-situ Raman on electrolyte: the O–H stretching region (3,000–3,700 cm⁻¹) shifts when water molecules coordinate to Mg²⁺ vs. remaining as free water. Tracks WIS regime stability over time. Also detects CN– stretching (~2,100 cm⁻¹) from any ferricyanide leaching from the cathode framework.
Copper Accumulation Tracking

ICP copper data is the single most important degradation metric. Plot Cu concentration (ppm) vs. cycle number for each C-rate. Extrapolate to determine the cycle count at which Cu exceeds the remediation trigger threshold (target: define this threshold during Phase 1). If Cu leach rate at C/4 is <5 ppm per 100 cycles, remediation intervals of 6+ months are feasible. If >50 ppm per 100 cycles, the copper contamination problem may dominate the system economics.

This data directly feeds the Rosemount continuous monitoring + electrowinning remediation loop described in the full proposal.

Statistical Design

Minimum n = 3 cells per condition (C-rate × temperature × electrolyte track). All capacity retention and coulombic efficiency values reported as mean ± standard deviation. Batch-to-batch reproducibility validated by requiring <10% coefficient of variation (CV) in initial discharge capacity across ≥3 synthesis batches before committing cells to long-term cycling. If CV exceeds 10%, investigate synthesis parameters before proceeding.

Go/no-go gate criteria apply to the worst-performing cell in the group, not the mean. A single outlier with capacity retention <80% triggers investigation even if the group mean is >95%. The conservative interpretation protects Phase 2 from survivorship bias.

Byproduct & Recovery Strategy

Phase 1 commits to a byproduct philosophy: every output stream is either a saleable product, a regenerable consumable, or processed on-site to non-toxic discharge. Zero stream leaves the lab in a form that creates downstream liability. The lab pays its own electric bill (load bank), recovers its own copper, and ships out a higher purity electrolyte than it received.

Revenue Streams
OutputYield & PurityRecovery Path
Recovered copper
From cycled-cell cathode leaching
5–15 g per 100 cycled cells. Dendritic powder, ≥99.5% purity after electrowinning.emew or ElectraMet bench rig in Zone A. Output sold to scrap refiner at LME spot (~$13–14/kg as of June 2026); at Phase 2 scale, routed back to Emerson sensor manufacturing as raw stock.
Cycled-cell electrodes
Postmortem residuals
Sub-gram Cu/Fe per cell, but stoichiometrically valuable.Phase 1: archive intact for forensic XPS/SEM/EDS. Phase 2: acid leach + electrowinning recovers active material. Zero landfill.
Regenerable Consumables
StreamRegenerationNet Consumption
Spent zeolite
Track B Ca/Sr removal column
NaCl brine elution — in-place, no removal from skid.≥50 regeneration cycles per bed. End-of-life: ground zeolite is non-toxic mineral filler, sold to construction-aggregate market.
NMP, DMSO, DMF
Synthesis / electrode wash solvents
Small distillation still in Zone B. Recovery: NMP ~95%, DMSO ~92%, DMF ~90%.~5–10% of nominal usage leaves as residue. Phase 2 goal: replace NMP with water-based binders (CMC/SBR, proven in Li-ion industry) to eliminate the stream entirely.
On-Site Detox — not waste, treatment
InputTreatmentDischarge
Ferricyanide wash waterNaOCl oxidation in-line. CN⁻ → cyanate (CN⁻O).Verified <1 ppm CN⁻ before drain. Cyanate is non-toxic at the expected concentration; municipal-permit compliant.
Copper-contaminated electrolyteInline ICP. <25 ppm Cu → routed to recovery loop above. >25 ppm Cu → electrowinning recovers ~95% before discharge.Discharge stream: <1 ppm Cu, pH 6.5–7.5, mineral salt water. Compliant with municipal wastewater permits in all 50 states.
Zero-Discharge Vision · Phase 2 target

NMP → eliminated (water-based binder substitution).
DMSO / DMF → 99% recovery via second-pass distillation.
Copper → 100% recovered, sold or reused.
Cells → recovered to active material at scale.

By Phase 2, the only outputs that leave the building are electricity (load bank discharge to Shell Space LEDs), recovered copper (sold), and mineral salt water (compliant discharge). Nothing else.

Stage 1 capex now itemized in companion proposal v14: $24–48K Zero-Discharge Capital line covers solvent distillation still (Zone B), NaOCl in-line dosing system (Zone A), bench-scale electrowinning rig (emew or ElectraMet contract), recovery-loop plumbing & DeltaV integration, and a Phase 2 lifecycle-assessment scope study. Phase 2 commits to publishing a peer-reviewed LCA demonstrating the closed loop.

Energy Budget — 9-Month Stage 1 Operation

The Phase 1 lab has a measurable, finite electricity demand that we should size honestly before we promise a load-bank offset. Equipment-by-equipment continuous and intermittent loads, the operating-hour assumptions, and the projected cost are below.

EquipmentContinuous (kW)Peak (kW)Hours/daykWh/day (avg)
NI HPS-17000 cycler (cycling load)51506–8 (active)120–240
Glovebox (Ar circulation, gas purification, lighting)3–452472–96
Environmental chamber (-40 to +85°C, intermittent)0.5412 (avg)12–30
HVAC + fume-hood makeup air (lab allocation, MN climate)10–151824240–360
Vacuum oven, ICP-OES, EQCM (intermittent)0.554–6 (avg)12–25
Solvent still + electrowinning rig (recovery loop, intermittent)0.254 (avg)12–20
Lighting + DeltaV controllers + computers3–452472–96
Daily total (mixed duty)~25~80–100540–870
9-Month Stage 1 Total
160–240 MWh
600–900 kWh/day × 270 days. Range reflects cycling intensity (number of parallel cell builds active).
Stage 1 Electricity Cost
$10K–$21K
MN industrial rate $0.07–0.10/kWh. Net of load-bank offset (~30–40% of cycling discharge feeds Shell Space LEDs).
Annual Run-Rate · Post Stage 1

After build-out, the lab’s steady-state electricity is approximately 200–320 MWh/year at $15K–$28K/year. Load bank discharge is the primary offset in Phase 1 (returns lighting and base load to Shell Space). At Phase 2 scale, the load bank ties into the building electrical sub-panel; a 10 MWh facility-scale aqueous BESS at 80% round-trip efficiency would offset roughly half of annual lab consumption while serving as the demonstration unit. The lab does not pay for its own electricity at Phase 1 scale; it does at Phase 2 and beyond.

13

Procedure I: Track B Purification Protocol

Convert industrial-grade MgCl₂ from Intrepid Potash (~95–97% purity) to battery-grade (≥99.9%) using the DeltaV-controlled purification skid. This process, once validated, becomes Product 6.

5-Stage DeltaV Purification Skid

1
Dissolve
2
Precipitate
3
Ion Exchange
4
Recrystallize
5
Final QA
01
Stage 1: Dissolution
Dissolve 5 kg industrial MgCl₂·6H₂O (Intrepid Potash) in 6 L heated deionized water (50°C) to produce a saturated brine (~5–6M). Stir 30 min. Solution will be yellow-brown due to iron/organic impurities. Monitor conductivity (Rosemount 228) and density (Micro Motion ELITE) continuously during dissolution.
02
Stage 2: Selective Precipitation
Add 2M NaOH dropwise via DeltaV-controlled Fisher valve at 5 mL/min while monitoring pH (Rosemount 3900). At pH 9.0–9.5: Fe(OH)₃, Ca(OH)₂, and heavy metal hydroxides precipitate while Mg remains in solution (Mg(OH)₂ precipitates at pH >10.5). DeltaV auto-stops NaOH feed when pH reaches 9.5. Stir 1 hour to complete precipitation. Filter through 1 µm cartridge.

This is the critical selectivity step. The pH window between Ca(OH)₂ precipitation (~pH 9) and Mg(OH)₂ precipitation (~pH 10.5) gives ~1 pH unit of operating margin.
03
Stage 3: Ion Exchange
Pass filtered brine through a column of Type 4A zeolite (bed volume: 5 L, flow rate: 2 BV/hr via DeltaV flow control). The zeolite selectively captures Ca²⁺, Sr²⁺, and remaining Fe³⁺ while passing Mg²⁺. Monitor outlet conductivity. When Ca < 1 ppm by inline probe, switch to collection. Regenerate column with 10% NaCl brine (10 BV) between runs.
04
Stage 4: Recrystallization
Evaporate purified brine under vacuum (70°C, 200 mbar) until supersaturation. Slow cool at 0.5°C/min to 10°C. MgCl₂·6H₂O crystallizes as large, colorless, transparent crystals. Collect by filtration. Mother liquor recycled to Stage 1. Expected yield: 60–75% per pass.
05
Stage 5: Final QA
Dissolve recrystallized product to 4.0M and run full QA suite (Procedure A, Steps 02–06): conductivity, pH, density, ICP-OES. Compare all values to Nedmag baseline within ±5%. Record as LOT-B-XXXX.

Track B electrolyte is NOT used in cells until ICP purity matches Nedmag baseline. Target convergence: Month 6–9.

Cost Ladder

StageInputPurityCost ($/ton)
Raw feedstockIntrepid Potash~95–97%$200–600
After precipitation + IXStage 3 output~99.5%+$400–600 processing
After recrystallizationStage 4 output≥99.9%+$400–900 processing
Battery-grade total≥99.9%$1,000–2,100
Nedmag importedNetherlands≥99.9%$1,500–3,000
14

Go/No-Go Gates & Milestone Criteria

Four checkpoints. What earns the next quarter of funding, and what kills the project.

M3
MONTH 3
First cells assembled and cycling

Track A: CuFe-PBA cathode synthesized via coprecipitation, characterized (XRD, TGA, ICP, BET). COP anode synthesized via solvothermal route, characterized (FTIR, TGA, BET). First prismatic cells built with Nedmag electrolyte. Initial cycling at C/10 and C/4 underway on NI HPS-17000. DeltaV skid commissioned for Track B. Formation data on ≥20 coin cells.

Gate Criteria

PASSCells cycle without catastrophic failure. CE >98% post-formation. pH within 4.91–7.02. Reproducible synthesis (≥3 batches, <10% capacity variance).
KILLNo cell survives 50 cycles. CE <90%. pH excursion >2 units. Synthesis irreproducible.
M6
MONTH 6
1,000+ cycles at grid rate. Degradation curve visible.
Primary Go/No-Go Decision Point

Track A: EIS impedance data reveals degradation trajectory. Copper accumulation rate measured via ICP at 500 and 1,000 cycles. Coulombic efficiency trend established. dQ/dV analysis shows whether redox peaks are shifting. Rate capability check confirms recoverable capacity. Track B: first purification runs complete, ICP on purified product.

Gate Criteria

PASSCapacity retention >95% at 1,000 cycles (C/4). Cu leach rate supports remediation interval >6 months. EIS impedance growth <20%. CE >99.5%.
CONDITIONALRetention 80–95%. Cu leach 5–50 ppm/100 cycles. Continue with modified protocol and monthly reviews.
KILLRetention <80% at 1,000 cycles. Cu >50 ppm/100 cycles. Impedance growth >50%. Chemistry does not scale to grid rates.
M9
MONTH 9
Track B electrolyte tested against Nedmag baseline

Track B: ICP analysis compares domestically purified MgCl₂ against Nedmag spec across all measured impurities. If purity matches (≥99.9%, all impurities within 5% of Nedmag levels), cells switch to domestic electrolyte for remaining cycling. Purification process recipe documented for Product 6 development. Parallel: Track A cells at 2,000–4,000+ cycles depending on rate.

Gate Criteria

PASSTrack B purity ≥99.9%. Cell performance on domestic electrolyte matches Nedmag baseline within 5%. Purification process documented and reproducible (≥3 batches).
CONDITIONALPurity 99.5–99.9%. Performance gap 5–15%. Additional purification step required (second recrystallization or activated carbon polish).
KILL (Track B only)Cannot reach 99.5% purity. Performance gap >15%. Track B abandoned; continue Track A with imported Nedmag. Product 6 thesis not viable for Intrepid Potash feedstock.
M12
MONTH 12
Phase 1 Report. Phase 2 Go/No-Go.

Everything on paper. Comprehensive dataset: cycle life projection at C/4 and C/2, copper remediation plan with economics, electrolyte reuse data, energy efficiency at each rate, domestic supply chain validation, temperature sensitivity map, electrode manufacturing reproducibility. Phase 2 decision made with data, not projections.

Deliverable

Phase 1 dataset + Phase 2 recommendation: fund it, modify it, or stop. The dataset includes: ≥3,000 cycles at C/2 with full characterization, degradation model with EOL projection, copper management protocol, domestic electrolyte qualification status, Bill of Materials update with real costs, and preliminary cell design for Phase 2 scale-up.

15

Twelve-Month Timeline

M1
M2
M3
M4
M5
M6
M7
M8
M9
M10
M11
M12
A: Synthesis
A: Formation
A: Cycling (C/4, C/2)
A: EIS/ICP/pH
A: Thermal study
B: DeltaV commission
B: Purification dev
B: Cell qualification
Reports & Gates
Synthesis Setup / Formation Cycling Characterization / Qualification Thermal Study Go/No-Go Gate
16

Data Management & Reporting

Data Architecture

Continuous (DeltaV Historian)

• Rosemount 228 conductivity (1 Hz)
• Rosemount 3900 pH/ORP (1 Hz)
• Rosemount 372 dissolved O₂ (0.1 Hz)
• Rosemount 928 H₂ gas (1 Hz)
• Micro Motion ELITE density (1 Hz)
• Fisher valve positions (1 Hz)
• Skid temperatures, pressures (1 Hz)

All continuous data logged to DeltaV historian with 1-year retention minimum.

Periodic (EECOMOBILITY + Manual)

• HPS-17000 cycling data (1 Hz per cell)
• EIS Nyquist/Bode plots (per session)
• ICP-OES elemental analysis (per sample)
• XRD diffractograms (per postmortem)
• SEM/EDX images (per postmortem)
• Electrode mass/thickness logs
• Batch synthesis records

EECOMOBILITY handles automated analysis. Manual entries in structured lab notebook (electronic).

Reporting Cadence

FrequencyAudienceContent
WeeklyLab teamCycling status dashboard (EECOMOBILITY), anomaly flags, equipment status, upcoming characterization
MonthlyProject sponsorCapacity retention trends, copper accumulation curve, Track B progress, budget vs. actual, risk register update
Quarterly (M3, M6, M9, M12)Executive / Go/No-Go committeeComprehensive milestone report with gate criteria assessment, go/no-go recommendation, Phase 2 implications

Intellectual Property

The published paper (Chen et al.) places the fundamental chemistry in the public domain. Emerson’s IP position derives from process know-how, not formula exclusivity: NI test protocols for cycling and degradation characterization, the copper remediation loop (Rosemount monitoring + electrowinning recovery), the DeltaV purification recipe, and aqueous-specific BMS algorithms for Zitara. All lab notebooks, synthesis records, and characterization data are Emerson trade secrets. External publications require executive approval.

Salt. Water. Iron. Copper.
Now prove it.
Connor Scanlan · Material Attendant · Shakopee · 2026