Random and impact copolymer polypropylene run the same reactor train up to one decision. Either ethylene gets co-fed into the main loop, or a second gas-phase reactor is added downstream to grow a rubber phase inside the finished homopolymer.
That one fork is the difference between a clear pipe resin and a sub-zero-tough automotive grade. Everything before it is shared chemistry; everything after it is the same finishing line.
The sequence below runs from the propylene feed onward, naming the resin property and grade slot each step delivers. Knowing where a property is set lets you sanity-check a COA instead of trusting the spec sheet blindly.
Step 1 — The Propylene Feed and the Ziegler-Natta Catalyst System
Copolymer PP production starts with polymer-grade propylene (typically ≥99.5% purity) and a supported Ziegler-Natta catalyst. The catalyst sets stereochemistry before a single chain has grown — it decides isotacticity, which is what makes commercial PP rigid at all.
The modern industrial system is TiCl₄ supported on MgCl₂, activated by triethylaluminum cocatalyst. Two donors do the fine-tuning: an internal donor (diisobutyl phthalate) and an external alkoxysilane such as cyclohexylmethyldimethoxysilane.
The donors are not optional dressing. They control how cleanly the chain stays isotactic and how evenly ethylene incorporates on a copolymer route.
Commercial isotactic PP lands at an isotactic index of 85-95%, a direct output of the donor package. A metallocene single-site catalyst is the narrower-distribution alternative, but sees little commodity use.
Feed purity is the quiet constraint. Ziegler-Natta sites are poisoned by trace moisture, oxygen, and sulfur, so the propylene is dried and guard-bed treated before the reactor.
Step 2 — Primary Polymerization in the Bulk Loop Reactor
Primary polymerization happens in a liquid-pool loop reactor running liquid propylene as its own solvent at roughly 60-80 °C and 30-40 atm. This is where the isotactic homopolymer matrix that every copolymer route shares gets built.
Chains grow off the active TiCl₄ sites as the mixture circulates. Hydrogen is dosed in as the chain-transfer agent: more hydrogen means shorter chains, which means higher melt flow.
That hydrogen dose is the single lever setting the grade’s melt flow index target, dialed in right here in the loop.
Commercial trains run two loops in series. The bulk-phase route caps ethylene at roughly 5 wt%, so any grade needing more comonomer or a rubber phase moves downstream — where the fork begins.
Run propylene alone through these loops and the output is homopolymer polypropylene, the rigid baseline that melts at 160-171 °C. Both copolymer routes branch from this same loop output; what changes is whether ethylene enters, and when.
Step 3 — The Random-vs-Impact Fork
The random-vs-impact decision is one physical choice on the train. Co-feed ethylene into the loops for a random copolymer, or hold the loop matrix and add a second gas-phase reactor to grow a rubber phase for an impact copolymer.
Same upstream loops, two different resins. The taxonomy of what copolymer polypropylene is sits in the definitional companion; the focus here is where the routes diverge.
| Property outcome | Random route | Impact route |
|---|---|---|
| Ethylene entry | Loops, 2.7-8 wt% | Second reactor as rubber |
| Reactors used | Two loops only | Loops + gas-phase reactor |
| Melting point | ~135-150 °C | Matrix Tm retained |
| Clarity | High (clear) | Set by domain size |
| Cold-impact toughness | Modest | Several-fold higher |
Random Route — Ethylene Co-Fed Into the Loops
Feeding ethylene into the loops at 2.7-8 wt% inserts ethylene units randomly along the chain. That disrupts the isotactic sequence, so crystallinity — and the melting point — drops, trading stiffness for clarity.
A measured Spheripol random grade at 2.7-3.0 wt% ethylene sits at Tm 144.6 °C, well below homopolymer’s 160-171 °C, with 4.8 wt% xylene-soluble fraction. It is made in two loops only, with no gas-phase reactor.
PetroChina Dushanzi T4401 is the real-world output of this loop-reactor random route. The PP-R pipe grade runs MFR 0.25 g/10 min (230 °C/2.16 kg, ASTM D1238/ISO 1133), density 0.90 g/cm³, rated to 95 °C hot-water service.
One spec-sheet sanity-check: some vendor pages list T4401 at 164-170 °C, which is homopolymer range. A genuine random copolymer cannot melt that high — random ethylene insertion physically prevents it. Expect ~135-150 °C.
Impact Route — A Rubber Phase Grown in a Second Reactor
The impact route keeps the loop output as a rigid homopolymer matrix. It routes that matrix into a second gas-phase fluidized-bed reactor (70-80 °C, 25-35 bar), where ethylene-propylene rubber (EPR) grows in-situ as a dispersed phase at 5-15 wt%, sometimes up to ~25%. Those domains absorb impact energy the rigid matrix cannot.
“Block copolymer” is the common trade name, but it is a misnomer. The impact route makes no chain with covalently joined blocks — it makes a heterophasic alloy: discrete EPR rubber domains dispersed in an isotactic homopolymer matrix. That is why impact and stiffness tune almost independently, by rubber fraction rather than chain architecture.
Rubber-domain size is the clarity-versus-impact lever, set by EPR molecular weight relative to the matrix. Coarse ~1 μm domains scatter light to 98.8% haze (opaque) at 14.5 kJ/m² impact. Fine ~100 nm domains drop haze to 13.5% (nearly clear) while raising impact to 25.1 kJ/m².
Lower-MW EPR disperses into smaller domains — better clarity and toughness from the same process. This output lands in the impact copolymer (ICP) grade slot for sub-zero and drop-test applications.
Step 4 — Degassing, Compounding, and Pelletizing
After polymerization the powder still holds unreacted monomer, carries no stabilizer, and exists as fluff rather than pellets. The finishing chain fixes all three — the step most process explainers skip.
First the polymer flashes and degasses to strip and recover unreacted propylene (and ethylene, on copolymer grades) for recycle to the reactor. Skip clean degassing and residual monomer surfaces later as odor and voids.
Then comes additive compounding. A stabilizer package of antioxidants and acid scavengers gets melt-blended into the powder, along with any property modifiers the grade needs.
A clarifier belongs here, not in the reactor — the 12.5% haze on that random grade came from 3000 ppm of a sorbitol clarifier at the compounder, not the polymerization.
The melt is then pelletized into uniform pellets that feed a customer’s extruder or molder cleanly. Grade QC closes the loop: melt flow is measured by ASTM D1238 / ISO 1133 at 230 °C/2.16 kg. That number on the T4401 COA confirms the Step 2 hydrogen dose hit its target.
The Full Picture, Start to Finish
The lever that decides everything is small and early: whether ethylene enters in the loops or as a separate rubber phase downstream. One co-feed makes a clear, lower-Tm random resin; one extra reactor makes a cold-tough heterophasic alloy.
A COA also reads differently once you know the train. Tm, haze, impact, and MFI each trace back to a stage — comonomer in the loop, rubber-domain size in the second reactor, clarifier and hydrogen dose at the finish. Read them as process outputs, and an unreliable value like a 168 °C random-copolymer Tm gives itself away.