Industry Insights & Corporate News
I’ve been responsible for the tetrachloroethane production unit at this site for fifteen years. We run the acetylene chlorination route exclusively, giving us a 1,1,2,2-tetrachloroethane stream that doesn’t carry the 1,1,1,2-isomer baggage you get from redistilling chlorocarbon waste. Over those years I’ve seen formulators treat this solvent like a commodity—open a drum, reel off some liters, close it poorly—then call me to ask why their NBS bromination dropped selectivity or why their AlCl₃ complex started generating HCl at 40°C. This isn’t a generic chlorinated solvent. It’s a dense, thermally sensitive workhorse that will turn on you if you ignore moisture, light, and residence time at temperature. Our typical delivered batch purity sits at 99.85% by GC-FID (DB-624, 30m x 0.32mm, split injection), with a moisture median of 60 ppm by Karl Fischer coulometric titration. But those numbers only hold if you keep the system closed and nitrogen-blanketed. I’ve pulled samples from a customer’s day tank after a weekend of ambient breathing in a Houston July—moisture hit 340 ppm and acidity as HCl topped 12 ppm, easily enough to stall a Grignard initiation or turn a clean Friedel-Crafts into a tar pit. We batch-test every isotank and drum using ICP-MS for metals, coulometric KF for water, and potentiometric titration for acidity, not a color-card comparator. The table below is what we ship against, and what you need to know before the material ever touches your reactor.
| Property | Test Method | Specification | 12-Month Lot Average (n=87) |
|---|---|---|---|
| Assay (1,1,2,2-Tetrachloroethane) | GC-FID, DB-624 column | ≥99.5% | 99.86% |
| 1,1,1,2-Tetrachloroethane isomer | GC-FID, same column | ≤0.10% | 0.02% |
| Moisture | Karl Fischer coulometric | ≤100 ppm | 58 ppm |
| Acidity (as HCl) | Potentiometric titration | ≤5 ppm | 1.8 ppm |
| Non-volatile residue | Gravimetric, 105°C, 2h | ≤10 ppm | 3 ppm |
| Iron (Fe) | ICP-MS | ≤0.5 ppm | 0.12 ppm |
| Color (APHA) | Visual comparison | ≤10 | 5 |
| Specific gravity @25°C | Digital density meter | 1.585–1.595 | 1.592 |
| Distillation range (5–95 vol%) | ASTM D1078 | 145.0–147.0°C | 146.2–146.8°C |
We internally tighten every spec that relates to downstream catalytic performance. The iron number isn’t a regulatory box to tick; we learned from a polycondensation customer whose ytterbium triflate catalyst deactivated irreversibly at 0.8 ppm Fe, so we keep it below 0.5 ppm by using glass-lined distillation and storage. The low isomer content—our isomer runs 0.02% versus the industry typical 0.2%—comes from a process choice I’ll get into later. It matters when your reaction is stereochemically sensitive, like a chiral auxiliary synthesis where the 1,1,1,2-isomer solvates diastereomeric salts differently and shifts your ee by a point. I’ve had that phone call.
This is the single biggest use we serve, and where I’ve got the most battle scars to share. In a typical radical benzylic bromination—say, converting 4-methylbiphenyl to its bromomethyl derivative—the reaction runs at reflux, and the boiling point of tetrachloroethane (146°C) puts your jacket temperature right around 147–150°C with good reflux. The practical dosage we see most formulators land on is 5–7 volumes of solvent per mass of substrate, not because of solubility limits but because that volume gives a manageable slurry of NBS. NBS is only sparingly soluble; you’re running a suspension. A retreat-curve impeller in a glass-lined reactor running at 150 rpm keeps the solid NBS from settling on the bottom where local hotspots can kick off ionic side reactions. If you try to go lower volume, say 3 volumes, the slurry becomes so thick that you start seeing heating lag and then overshoot on the jacket, and I’ve witnessed a 12°C temperature spike in a 2000-L vessel that generated enough HCl to etch the sight glass seal. The solvent flash point is effectively nil—non-flammable—but the decomposition products are corrosive and will find any weak point in your glass lining or gaskets.
Exactly how you introduce tetrachloroethane into the process matters more than most engineers think. We recommend pumping it from a nitrogen-padded ISO container via a magnetic-drive gear pump with a 1-m positive suction head—because specific gravity 1.59 will vapor-lock a pump sized for 1.0 SG at zero NPSH, even if the pump curve says it’s fine. I’ve seen a C-face pump cavitate after 20 minutes, stop delivering, and then the operator overrode the low-flow alarm and ran the pump dry, sending metal shavings into the line. Use a Coriolis mass flow meter downstream, not a volumentric meter; density shifts from temperature affect volume measurement by 0.1% per °C, and that error adds up when you’re charging a precise solvent ratio for kinetics. Charge the substrate solids and NBS first under nitrogen, then add the solvent via a dip tube that reaches below the liquid surface. Opening a manway and dumping 200 L from a drum under ambient air will inject enough moisture to raise your bulk water from our 60 ppm to over 150 ppm within the headspace, and radical initiators like AIBN don’t forgive that.
Moisture is the silent killer. In an NBS bromination, water at levels above 200 ppm will hydrolyze NBS into HOBr and HBr, and then the HBr can catalyze ionic bromination on the aromatic ring, dropping your benzylic selectivity from 95% to below 80% in a matter of minutes at reflux. We tell customers: if you’re going to hold an open drum for more than a shift, dry the solvent over 3A molecular sieves for 24 hours at room temperature, or install a closed nitrogen loop with a drier cartridge. Our delivered moisture is low, but it’s your job to keep it low. One facility tried to solve this by adding a long nitrogen sparge before heating—that stripped solvent and cooled the liquid so much that it took an extra two hours to reach 110°C, and they still got emulsions in the workup because the HBr formed during that slow heat-up attacked the glass lining and pulled silica into the phase. The order of operations matters:
Light is another vector. Tetrachloroethane left in clear glass sight lines or borosilicate piping under UV-rich plant lighting will generate acidic decomposition at about 2 ppm HCl per day. I’ve quantified it. Trace the carbon-steel swivel joints on a tote line after six months of Florida sun and you’ll find pitting that matches that rate. Always use amber or black-pigged drums, black iron piping, and if you have sight glasses, sleeve them with UV-blocking film. That’s a five-dollar fix that saves you a thousand-dollar catalyst repurchase.
One final note on this application: if you’re running a photobromination with a mercury lamp, the solvent transparency is fine down to 300 nm, but the decomposition path accelerates, and you’ll generate enough HCl within four hours to shift pH and quench the radical chain. We recommend an inline quartz UV filter to strip the wavelengths below 330 nm, which cuts the acid formation by 70% without slowing the bromination significantly. We picked that number after a customer set up a continuous photochemical reactor and saw yield drop from 88% to 71% when they switched from a Pyrex-encased lamp to an open quartz one. Our tetrachloroethane’s low iron and low photolabile impurities make this effect more predictable than with mixed-isomer solvents, but it’s still physics, and I can’t beat physics.
When a pharmaceutical intermediate requires an aromatic acylation with aluminum chloride, tetrachloroethane often replaces carbon disulfide because it gives a higher operating temperature without the fire risk. The typical ratio is 2–3 liters of solvent per mole of acyl chloride substrate, and the AlCl₃ is first slurried in a portion of the solvent at 0–5°C before the substrate is added. If you let the slurry temperature creep above 25°C during the complex formation, even for ten minutes, the Lewis acid will begin to abstract chlorine from the solvent, liberating HCl and forming a dark, tar-like aluminum chloro-complex. The exotherm can be brutal. I’ve reviewed a batch that self-heated from 5°C to 62°C in a 3000 L glass-lined reactor when the jacket brine circulation failed and the operators didn’t catch it for twenty minutes. The subsequent acylation gave 32% yield—with 71% of the recovered solvent turning black and non-recoverable. That batch was years ago, but I still keep the reactor temperature chart pinned in my office.
The integration point is critical: the tetrachloroethane must be pre-cooled to 0°C and dosed through a jacketed static mixer just ahead of the reactor inlet. This prevents a localized warm spot where the new solvent meets the stirred slurry. We’ve had customers successfully run this on an anhydrous skid with a flow ratio controller that modulates the solvent pump speed based on the slurry’s temperature at the reactor inlet. Anything less than a RTD sensor with a five-second response time is asking for trouble. And do not—I repeat, do not—use a centrifugal pump to transfer this cold solvent without ensuring the mechanical seal is compatible with trace acidity; a single-face carbon vs. ceramic seal will eventually etch, and then you’ll have air in-leakage that brings moisture into your AlCl₃ complex. Double mechanical seals with a barrier fluid are standard at our plant for this reason, and if your skid doesn’t have one, stop and rethink.
We told a customer recently that tetrachloroethane cannot be used in Friedel-Crafts reactions with zinc chloride at temperatures above 130°C. Their R&D had a lab-scale success, but on scale-up, the zinc chloride catalyzed dehydrochlorination to 1,1-dichloroethylene and HCl at an alarming rate above 125°C, popping the rupture disk on a 500-gallon vessel. That’s not a solvent limitation—it’s a reactive hazard that the literature ignores. We’ve since added a note to our specification sheet: “Not recommended as a reaction medium for Lewis acid-catalyzed processes above 120°C unless decomposition byproducts are managed.” So if you’re exploring something frontier, call me and we’ll run a DSC screen on your catalyst cocktail to see where the exotherm lives.
Its density of 1.59 g/cm³ makes tetrachloroethane a useful wash solvent for extracting high-value polar products from aqueous streams when you want the heavy phase to be the organic. A fine chemical manufacturer we supply uses it in a Karr reciprocating-plate extraction column to recover a cyclopropane carboxylate intermediate from a 20% sodium chloride brine. The organic-to-aqueous ratio runs at 0.8:1 by volume, and the column operates with the heavy phase as the continuous phase, which means the dispersed aqueous droplets rise upward. If your column has been designed for a solvent of SG 1.1, you’ll need to re-check the plate spacing and the drive amplitude—the higher interfacial tension against the brine will give you droplet sizes closer to 1–2 mm instead of 0.5 mm, and the specific throughput per cross-sectional area will drop by 30%. We’ve lab-tested it. That same customer also learned that trace surfactants (below 0.1% of a quaternary ammonium phase-transfer catalyst) can invert the phases at the column’s bottom coalescer, sending their product down the brine outlet leg. A coalescing filter element downstream of the column, sized for the heavier phase, saved them from a week of rework.
In a simple batch extraction in a stirred vessel, our solvent will separate cleanly from water within 8 minutes, but only if the aqueous phase hasn’t been loaded with more than 3% methanol or acetonitrile from a previous reaction step. More than that and you’ll form a rag layer that takes two hours to fully settle, even with brine. We tell our customers to strip the aqueous phase to below 2% organics before the extraction, or use a multistage countercurrent setup with raffinate recirculation. And yes, the solvent hydrolyzes slowly in neutral water—our long-term data show about 3 ppm HCl generated per week at 25°C when saturated with water—so don’t store the recovered solvent wet for a shutdown cycle. Strip it and dry it immediately or you’ll corrode your recovery still.
If you’re using a centrifugal extractor, like a Podbielniak, the sump must be explosion-proof even though tetrachloroethane is non-flammable. The rationale isn’t for the solvent itself; it’s because a mechanical failure in the bearings can generate enough heat to decompose the solvent, and the HCl vapor released can then stress-crack stainless steel. We mandate an inert nitrogen purge of 3 L/min in the housing for every 1000 L/h throughput based on a root cause analysis we did after a customer’s bowl seizure. The purge also keeps moisture out of the solvent film on the internals.
I already mentioned the wiped-film evaporator. Fifteen years ago we finished tetrachloroethane with a conventional batch distillation column operating at 135°C bottom temperature, 4-hour residence time at that heat. That process generated 250–500 ppm tetrachloroethylene and 50–100 ppm trichloroethylene from thermal dehydrochlorination. Those impurities didn’t budge our COA assay by much, so we thought we were fine—until a Japanese pharmaceutical firm doing a palladium-catalyzed coupling traced a 1.2% yield loss across a six-month campaign to tetrachloroethylene poisoning the zero-valent palladium complex. They proved it by spiking fresh solvent with the same level of tetrachloroethylene and watching the turnover number drop proportionally. That was a three-million-ton problem in a 2000 L reactor. So we tore out the batch still and installed a continuous wiped-film unit that operates at 120°C wall temperature and 50 mbar absolute, with a residence time of 12–15 seconds. The decomposition byproducts now run <30 ppm tetrachloroethylene and <10 ppm trichloroethylene. Not zero, but low enough that the catalyst sees effectively pure solvent. That wiped-film unit runs in a glass-lined alloy 20 housing, further keeping metal leaching to nothing. I won’t pretend it was cheap, but the process improvement paid back in eighteen months just from reduced customer rejections. And it’s something a competitor will never capture if they’re still boiling the pot for hours. When a customer’s QA sends us an OOS, the first thing I ask is whether their parallel run with another supplier’s material showed the same impurity peaks. Often, it did, and they just never looked.
Another item we do that isn’t in the standard playbook: we ship every bulk container with a headspace gas composition sticker that lists O₂, N₂, and moisture content as measured by a sampling port integrated into the manway gasket. That tells the receiving engineer whether the blanket held during transit. If they see O₂ above 1% or dewpoint above -20°C, they reject the tote without opening it. We decided to do that after a customer’s receiving area left an opened container in the rain for a weekend and then blamed our purity when the moisture spiked. I’ve got no patience for that, but the sticker makes the point before the argument starts.
Air is the enemy. Light is the accomplice. Keep both out, or don’t blame the solvent.