Vulcanization

Sticky tangle to elastic network

Pick up a piece of raw, unvulcanized natural rubber and it behaves like cold chewing gum. It's tacky, it sags under its own weight in summer heat, it cracks and turns brittle in winter cold, and if you stretch it, it stays stretched. At the molecular level it is nothing but an enormous tangle of long polyisoprene chains — each maybe 5,000 to 20,000 isoprene units long — held together by weak van der Waals forces. Because nothing covalently connects one chain to its neighbor, the chains slide past each other and the whole mass flows. That's why pre-1840 rubber goods were commercial disasters: Charles Goodyear's early rubber mailbags and life preservers melted into stinking puddles on hot days.

Vulcanization fixes this by stitching the chains together. When you heat rubber with sulfur, the sulfur builds covalent bridges — short chains of one to about eight sulfur atoms — that span from one polymer chain to another. After enough of these cross-links form, the sample is no longer a collection of independent chains. It is a single, sample-spanning molecule: one giant covalent network. Now when you stretch it, the cross-links act like anchors. The segments between anchors uncoil and store elastic energy, and when you let go, entropy snaps them back. Sticky flow becomes springy recovery. That transition — from a viscous thermoplastic to an elastic thermoset — is the entire point.

The remarkable part is the leverage. You don't need much sulfur. A typical soft-rubber recipe uses just 1–3 parts of sulfur per hundred parts of rubber by mass (the rubber industry's "phr" unit). That works out to roughly one cross-link for every few hundred isoprene units. A change of a percent or two in composition flips the material from useless goo to engineering material.

The chemistry: sulfur bridges at the double bond

Natural rubber is cis-1,4-polyisoprene. Each repeat unit carries exactly one carbon–carbon double bond, and it is that double bond — and the allylic C–H bonds right next to it — that sulfur attacks. (This is why a fully saturated polymer like polyethylene cannot be sulfur-vulcanized at all; with no double bonds, there is no reactive site, and you must use peroxide cross-linking instead.)

The repeat unit and the cross-linking reaction, schematically:

cis-1,4-polyisoprene repeat unit:

        CH3
         |
  ~CH2 - C = CH - CH2~      (one C=C per isoprene unit)
              ↑
         allylic positions: where sulfur acts

Cross-linking (overall):

  chain A ~CH2-C(CH3)=CH-CH2~        chain A ~CH2-C(CH3)=CH-CH~
                              + Sx →                      |
  chain B ~CH2-C(CH3)=CH-CH2~                            Sx   (a -C-Sx-C- bridge)
                                                          |
                                     chain B ~CH2-C(CH3)=CH-CH~

Elemental sulfur exists as eight-membered crown rings, S₈. On its own that ring is unreactive at rubber-processing temperatures, so unaccelerated vulcanization (Goodyear's original recipe) is slow and inefficient — it can take hours at 140 °C and burns a lot of sulfur into wasteful intramolecular chains and dangling ends that never connect two different chains. The accepted mechanism is a radical chain process at the high temperatures and short times of modern cures: the S₈ ring is opened into reactive sulfur diradicals or, in accelerated systems, into polysulfidic intermediates that abstract an allylic hydrogen, then add sulfur across to a second chain. The result is a distribution of bridges:

  -C-S-C-        monosulfidic   (x = 1)   thermally stable
  -C-S-S-C-      disulfidic     (x = 2)   moderate stability
  -C-Sx-C-       polysulfidic   (x = 3-8) flexible, but heat-labile

The S–S bond is the weak link. A typical S–S bond dissociation energy is only about 240–270 kJ/mol, versus roughly 270–310 kJ/mol for C–S and ~350 kJ/mol for C–C — and the S–S bonds inside a long polysulfide bridge are weaker still. So a long polysulfide bridge (–C–S₆–C–) is the first thing to break under heat or oxidation. That single fact drives almost all the engineering trade-offs in rubber: short monosulfidic links give heat resistance and low compression set, while long polysulfidic links give better fatigue and tear resistance. The cure system is tuned to bias the distribution.

Accelerators, zinc oxide, and the modern recipe

Nobody vulcanizes with sulfur alone anymore. The breakthrough that made the modern tire possible was accelerated sulfur vulcanization, introduced in the 1900s–1920s. A real recipe contains five ingredient classes, and the cure chemistry is a small dance between them:

Ingredient	Typical loading (phr)	Role
Rubber (e.g. NR/SBR)	100	The polymer backbone — defines "phr"
Sulfur	1–3 (soft) · 25–50 (ebonite)	The cross-linker
Accelerator (CBS, MBT, TBBS…)	0.5–2	Speeds and orders the reaction 100×+
Zinc oxide (ZnO)	3–5	Activator — forms the active zinc complex
Stearic acid	1–3	Solubilizes Zn²⁺; co-activator
Carbon black (filler)	30–60	Reinforcement (not part of cure, but vital)

The accelerator (a sulfenamide such as N-cyclohexyl-2-benzothiazolesulfenamide, CBS, or the parent thiazole MBT) reacts with sulfur and with the ZnO/stearic-acid pair to form a soluble zinc–accelerator–polysulfide complex. This complex is the true sulfurating agent: it delivers sulfur to the rubber's allylic sites in a controlled, low-temperature, fast reaction. The payoff is enormous — cure time drops from hours to a few minutes, the temperature needed drops, far less sulfur is wasted on useless intramolecular cyclic structures, and you gain a scorch delay: a built-in induction period that lets you shape and mould the compound safely before the network suddenly sets. Sulfenamide accelerators are prized precisely for that delayed-action safety.

Zinc oxide deserves a special mention because it appears in essentially every tire on Earth. The Zn²⁺ ion coordinates the accelerator and the growing polysulfide, holding the sulfur in a reactive geometry and helping cleave it into shorter, more efficient bridges. Stearic acid's job is to convert solid ZnO into a soluble zinc stearate so the zinc is mobile in the rubber matrix. Drop the ZnO and the same cure runs slower, hotter, and yields a weaker, less heat-resistant network.

What the numbers look like

Vulcanization is one of the clearest cases in materials chemistry where a tiny chemical change produces an order-of-magnitude property change. Approximate figures for natural rubber before and after a conventional cure:

Property	Raw rubber	Soft-vulcanized (≈2 phr S)	Ebonite (≈40 phr S)
Tensile strength	~2–3 MPa	~20–30 MPa	~50–70 MPa
Elongation at break	flows / no clean break	500–800 %	2–10 %
Young's modulus	very low; creeps	1–5 MPa	~2–4 GPa
Cross-link density	~0	~1 per 200–300 units	~1 per 5–10 units
Behavior on heating	softens, flows, melts	holds shape to ~200 °C	rigid; chars not melts
Recovery after stretch	permanent set	elastic snap-back	essentially inelastic

The elasticity itself is governed by rubber-elasticity theory, where the shear modulus is proportional to cross-link density:

G = ν · k_B · T      (or  G = (ρ / M_c) · R · T )

  ν   = number density of network chains (cross-links per volume)
  M_c = average molar mass between cross-links
  ρ   = density,  R = 8.314 J/(mol·K),  T = absolute temperature

Two non-obvious predictions fall straight out of this equation, and both are experimentally true. First, modulus rises with cross-link density (smaller M_c → stiffer rubber) — that's why ebonite is rigid. Second, and counterintuitively, the modulus rises with temperature: heat a stretched rubber band and it pulls back harder, because the restoring force is entropic, not energetic. A stretched, cured rubber band even releases heat when stretched and cools when relaxed (the Gough–Joule effect) — the opposite of a metal spring.

Reversion, over-cure, and the cure curve

You can't just leave rubber in the mould forever. Cross-link density does not climb monotonically; it peaks and then falls. Plot torque (a proxy for cross-link density) against time on a rheometer and you get the classic cure curve:

torque
  │                 ╭──────╮  ← optimum cure (t90)
  │                ╱        ╲___      reversion (NR at high T)
  │   scorch      ╱             ‾‾‾‾‾ marching cure (some SBR)
  │   delay   ___╱
  │ ________╱
  └────────────────────────────────→ time
   induction   cure        post-cure

The flat induction region is the scorch delay (safe processing time). Then cross-links form rapidly to an optimum, usually quoted as t90, the time to reach 90 % of maximum torque. For a tire compound at 150–160 °C that's on the order of 10–20 minutes; thin medical or balloon goods at higher temperature can cure in well under a minute. Push past the optimum and natural rubber shows reversion: the heat-labile polysulfide bridges break down faster than new ones form, cross-link density drops, and the part gets weaker and gummier. Choosing an "efficient vulcanization" (EV) system — high accelerator, low sulfur — biases the network toward stable monosulfidic links and largely eliminates reversion, at the cost of some fatigue resistance. This balance between cure time, reversion resistance, and final properties is the daily work of a rubber compounder.

Where vulcanized rubber shows up

• Tires. The dominant use by far — roughly 70 % of all rubber goes into tires. A passenger tire is a co-cured assembly of several different rubber compounds (tread, sidewall, inner liner, bead) each with its own sulfur and accelerator package, vulcanized together in a hot press at ~150–170 °C for ~10–15 minutes. The carbon-black reinforcement and the sulfur network together set the tread's wear, grip, and rolling resistance.
• Seals, gaskets, hoses, and O-rings. Anywhere a flexible part must hold a shape and snap back through millions of cycles. Low compression set — which demands stable monosulfidic cross-links from an EV cure — is the key spec.
• Synthetic elastomers. The same sulfur chemistry vulcanizes styrene-butadiene rubber (SBR, the workhorse of tire treads), polybutadiene, and nitrile rubber (NBR) — all of which keep the C=C double bonds sulfur needs. Saturated rubbers like EPDM keep just enough unsaturation (from a diene termonomer) to remain sulfur-curable.
• Ebonite / hard rubber. The high-sulfur extreme. Historically used for bowling balls, fountain-pen barrels, clarinet mouthpieces, and acid-resistant battery cases — wherever a cheap, machinable, chemically inert hard black solid was wanted before modern plastics.
• Footwear, conveyor belts, anti-vibration mounts, and erasers. The cheap, durable, grippy applications that quietly consume the rest of the world's vulcanized rubber.

Beyond sulfur: other cross-linking routes

	Sulfur vulcanization	Peroxide cure
Needs C=C double bonds?	Yes (allylic site)	No — works on saturated chains
Cross-link type	–C–Sx–C– (mono/di/poly)	Direct C–C bond
Cross-link strength	Weaker (S–S labile)	Strong, thermally stable
Heat / aging resistance	Good (EV) to moderate	Excellent
Flex / fatigue resistance	Excellent (polysulfide)	Lower — rigid links don't rearrange
Typical polymers	NR, SBR, NBR, EPDM	Silicone, EPM, EVA, saturated rubbers
Scorch safety / control	Excellent (with accelerators)	Trickier; oxygen inhibits surface cure

Other industrial cross-linkers fill specific niches: metal oxides cure chloroprene (neoprene); phenolic resins and quinoid systems give heat-resistant cures for butyl rubber; and radiation (high-energy electron beams) cross-links polyethylene and silicone without any chemical added. But for unsaturated general-purpose rubber, accelerated sulfur remains unbeatable on cost, control, and the prized combination of strength and flex.

Common misconceptions and pitfalls

• "Sulfur is a filler that bulks up the rubber." No — sulfur is a reactive cross-linker present at only 1–3 %. The bulk reinforcement comes from carbon black, an entirely separate ingredient. Confusing the two misses the whole point: it's the covalent network, not added mass, that creates the properties.
• "More sulfur always makes better rubber." Only up to a point. Beyond the optimum you get ebonite-like rigidity, and within a cure you can over-cure into reversion where natural rubber actually loses cross-links and weakens. The cure is optimized to a t90, not run to completion.
• "Vulcanized rubber can be melted and reshaped like plastic." It can't. Vulcanization makes a thermoset. The permanent covalent network is why a tire is durable and also why scrap tires are notoriously hard to recycle — you must break S–S bonds (devulcanization) rather than simply remelt.
• "You can sulfur-vulcanize any polymer." Only polymers with C=C double bonds (or allylic sites) react with sulfur. Polyethylene, silicone, and other saturated chains need peroxide or radiation cross-linking instead.
• "Zinc oxide and stearic acid are just fillers." They are the activator system. Without them, accelerated cure barely works; the soluble zinc–accelerator complex is what makes the sulfur reactive at low temperature and biases it toward efficient, shorter cross-links.
• "Vulcanization is melting the rubber." The opposite. Goodyear's stove accident worked because heat drove the cross-linking reaction, not because it melted anything. Heat is a reagent that supplies activation energy for sulfur to bond the chains; the product is the most heat-stable form of the material, not the most molten.