In the production of polymer-modified bitumen (PMB) for both paving and waterproofing membranes, the substitution of a raw styrene-butadiene rubber (SBR) emulsion by a pre-crosslinked SBR powder fundamentally changes the unit operations from demulsification and drying to dry-particle dispersion. Raw SBR emulsions, sometimes supplied at 30–50 wt% solids, introduce water that must be evaporated before or during blending, consuming approximately 2.26 MJ/kg of latent heat per kilogram of water and generating foaming if the heating rate is excessive. Crosslinked SBR powder, in contrast, contains no aqueous carrier but comprises vulcanized rubber particles with a gel fraction often exceeding 80% and a particle size distribution where the median diameter is commonly in the range 180–250 µm for coarse grades and 80–120 µm for fine grades used in high-performance binders. Because the crosslinked network prevents melting and flow, the powder is dispersed as discrete elastomeric inclusions in the bitumen matrix; this suspension differs thermodynamically and rheologically from a homogeneous polymer solution formed by thermoplastic elastomers or uncured SBR. The plant must therefore be configured to meter powder gravimetrically, protect the feed from moisture uptake, and pass the mixture through a high-shear dispersion zone before storage. Specifications applicable to such binders include EN 14023:2010 for PMB framework requirements and ASTM D5976-15 for type I SBR-modified asphalt used in roofing and waterproofing; for paving binders, the relevant grading framework is AASHTO M 332-22 or AASHTO M 320-21, with testing per AASHTO T 350-22, AASHTO T 316-19, ASTM D4402-15, ASTM D2872-12, and ASTM D6521-22.

What Limits Dispersion When a Non-Thermoplastic Crosslinked Powder Is Introduced at 180–190°C?

Mechanical dispersion of a non-thermoplastic powder in bitumen is governed by the balance between the cohesive strength of particle agglomerates and the hydrodynamic shear stress transmitted by the continuous phase. Crosslinked SBR particles cannot be dissolved or melted, so the process objectives are particle wetting, agglomerate size reduction, and uniform spatial distribution. The rotational viscosity of a paving-grade base bitumen measured by ASTM D4402-15 at 135°C typically falls between 300–500 mPa·s; at 185°C, the same bitumen may exhibit 80–150 mPa·s. This viscosity is sufficiently low for wetting but not always high enough to transmit extreme shear into agglomerates unless a colloid mill or multi-stage rotor-stator is used. Processing at 180–190°C represents a narrow control band because temperatures above 200°C accelerate thermo-oxidative aging and increase the release of volatile fractions, while temperatures below 175°C raise viscosity and reduce wetting. The critical processing window is therefore approximately ±5°C around 185°C when measured at the discharge of the high-shear device. Equipment capable of tip speeds of 15–25 m/s and gap settings of 0.25–0.50 mm is normally required; lower tip speeds may leave agglomerates above 100 µm that later settle. The specific mechanical energy input reported for stable dispersions is generally 0.8–1.5 kWh/t, although published data for a specific powder grade and plant configuration remains limited. The addition sequence also matters: the powder must be injected directly into the suction side of the external mill or into a high-shear zone rather than sprinkled onto the tank surface, because surface addition leads to floating agglomerates and long hydration times.

In a 20,000-litre vertical blend vessel equipped with a bottom-entering high-shear rotor-stator and an external colloid mill recirculation loop, the dry SBR powder is typically introduced through a side-stream eductor that combines bitumen at 185°C with metered powder at 4–10 wt% of the final batch. The powder hopper should be fitted with a loss-in-weight gravimetric feeder, a mechanical bridge-breaker, and a drying air or nitrogen purge, because the fine fraction below 75 µm can agglomerate under storage moisture and can also form a combustible dust cloud. Pre-drying in a fluid-bed dryer at 80–90°C for 2–4 h is required when the relative humidity of the storage area exceeds 60%; residual moisture above 0.5 wt% causes steam entrapment in the hot bitumen, visible foaming, and erratic density readings. The recirculation loop should provide at least 6–10 tank turnovers per hour so that the entire batch passes through the high-shear mill repeatedly; at lower recirculation rates, powder-rich regions remain near the feed point and create local viscosity inhomogeneities. On production lines, the most common failure modes are feeder belt slippage at high humidity, powder bridging over the feed screw when using very fine 80–120 µm powder, and excessive backpressure in the colloid mill if the gap is set below 0.20 mm. Combinations with amine-based anti-strip additives should be avoided during the dispersion stage because their polar functionality can migrate to the particle interface and alter wetting; if such additives are required, they are preferably dosed downstream after the dispersion loop.

Thermo-oxidative aging resistance of crosslinked SBR powder–bitumen blends under RTFOT and PAV protocols

Short-term aging is simulated by the rolling thin-film oven test per ASTM D2872-12 or AASHTO R 28-21 at 163°C for 85 min, and long-term aging is simulated by the pressure aging vessel per ASTM D6521-22 at 100°C under 2.07 MPa for 20 h. A crosslinked SBR powder, because its polymer network is already vulcanized, does not undergo the same chain-scission and molecular-weight reduction pathways as raw uncured SBR emulsion polymer during oxidative aging. Instead, aging primarily affects the bitumen matrix, concentrating polar asphaltenes and increasing stiffness. The powder can therefore act as a relatively inert reinforcing filler under aging, provided its interfacial region has been properly wetted. Retention of penetration after RTFOT is a commonly used specification criterion; some procurement documents require a minimum retained penetration of 55%, and properly dispersed blends with 4–8 wt% fine powder often exceed 60%, though published data for specific crosslinked SBR grades is limited and must be verified against ASTM D5-20. Softening point increase after RTFOT, measured by ASTM D36-14, is typically smaller for crosslinked powder systems than for raw emulsion systems at equal polymer content because no residual emulsifier or water-phase salts are present. However, long-term PAV aging may reveal a stiffening effect caused by the selective adsorption of maltenes onto the rubber particles; this can be detected by tracking the complex shear modulus master curve shift after PAV. The use of aromatic or naphthenic extender oils is therefore often necessary to maintain the maltene balance and prevent the aged powder-containing binder from becoming excessively brittle at low service temperatures.

Rheologically, a PMB made from crosslinked SBR powder behaves as a particle-filled viscoelastic liquid rather than a homogeneous polymer network. Its storage stability is governed by particle settling and flocculation, not by polymer phase separation alone. The EN 13399:2010 storage stability test requires samples to be held at 180°C for 72 h in a sealed tube, after which the softening point difference between the top and bottom thirds is determined by ASTM D36-14. For paving-grade binders, a difference of ≤2.5°C is often specified; for waterproofing membranes, some producers tighten this limit to ≤1.5°C. Crosslinked SBR powder systems can meet these limits only when the particle size distribution is controlled and the particle surface is sufficiently wetted; coarse grades above 250 µm may settle even at 180°C because the Stokes settling velocity scales with the square of particle diameter. Rotational viscosity measured by ASTM D4402-15 at 135°C should be monitored because an increase above 3.0 Pa·s often indicates either agglomeration or excessive bitumen oxidation during blending. Elastic recovery measured by ASTM D6084-21 at 25°C is a direct indicator of the rubber network contribution; values below 50% at 10°C elongation generally suggest incomplete dispersion or too low a powder addition level. The multiple stress creep recovery test per AASHTO T 350-22 provides additional evidence of the polymer contribution: non-recoverable creep compliance at 3.2 kPa should decrease relative to the unmodified binder, while the percent recovery should increase.

When a Colloid Mill Is Retrofitted for Dry Powder Injection at 180–190°C

Retrofitting an existing colloid mill for dry powder injection requires detailed attention to the powder feed location, rotor geometry, and temperature control strategy. The powder feed is best introduced into the bitumen stream through a high-shear eductor immediately upstream of the mill inlet; this arrangement ensures that each powder particle experiences the maximum shear field before it can float or reagglomerate. A colloid mill with a rotor tip speed of 15–25 m/s and a gap setting of 0.25–0.50 mm is considered suitable for medium-viscosity paving bitumen, but the actual gap must be adjusted depending on the particle size distribution of the powder. If the gap is too small, the backpressure on the feed pump increases and the mill may cavitate; if too large, the shear stress is insufficient to break agglomerates. The mill inlet and outlet should be equipped with pressure transmitters; an outlet pressure above 6 bar often indicates a blocked gap or insufficient pre-wetting. The recirculation loop should be sized to achieve 6–10 tank turnovers per hour, and the temperature at the mill discharge should be maintained at 180–190°C by jacket heating and, in longer loops, by external heat exchangers. A nitrogen blanket on the blending vessel and on the powder feed hopper reduces oxidative degradation of the bitumen and minimizes dust explosion risk. In a retrofit, the existing water-based emulsion feed pump is usually removed and replaced by a gravimetric feeder and a rotary airlock valve. The control system must be interlocked so that the powder feeder stops automatically if the mill recirculation flow rate drops below the minimum set point; otherwise, dry powder can compact in the mill and cause mechanical damage.

Performance grading of the modified binder should follow AASHTO M 332-22, with high-temperature continuous grade determined from the rutting parameter G*/sin δ before and after RTFOT and from the non-recoverable creep compliance Jnr after RTFOT. Crosslinked SBR powder tends to improve elastic recovery and reduce Jnr at high temperature, but the effect is smaller than that of linear SBS or high-molecular-weight uncured SBR because the crosslinked particles act as reinforcing filler rather than forming a continuous network. At 3.2 kPa stress, Jnr values of 0.5–1.5 kPa⁻¹ may be observed for 6–8 wt% powder-modified binders, whereas the base bitumen often exceeds 3.0 kPa⁻¹; published data for a given source and powder grade is necessary to set specification limits. Low-temperature performance is more sensitive to powder particle size and addition level. Fine powder with d50 below 100 µm can distribute stress more uniformly, while coarse powder above 250 µm can act as stress concentrators and reduce the critical cracking temperature measured by AASHTO T 313-22. Addition levels above 12 wt% typically produce rapid increases in low-temperature stiffness and may reduce the blend’s ductility below the minimum required by ASTM D113-17 at 25°C; addition levels below 4 wt% usually provide insufficient elastic recovery for dense-graded asphalt applications.

Phase morphology, gel fraction thresholds, and compatibility criteria

Phase morphology of crosslinked SBR powder-modified bitumen is characterized by discrete rubber particles dispersed in a continuous bitumen phase, with a size distribution that depends on the powder grinding method and the intensity of mixing. Fluorescence microscopy at 400× magnification typically reveals spherical or irregular particles with a bright yellow-green appearance against a darker bitumen background; image analysis can quantify the equivalent circle diameter and the total area fraction of rubber. A gel fraction threshold above 80% is important because a partially cured powder may swell excessively in bitumen, increase binder viscosity, and produce uneven phase distribution. The swelling ratio in toluene, measured by immersion at 23°C for 72 h, is often 3–5 for properly vulcanized SBR powder; significant deviation indicates either under-curing or excessive filler. The asphaltene and maltene balance of the base bitumen is critical: highly compatible binders with a low asphaltene fraction wet the nonpolar SBR powder surface more readily, while highly asphaltenic binders may require the addition of 5–10 wt% aromatic extender oil to improve interfacial adhesion. The glass transition temperature of the crosslinked SBR phase, measured by differential scanning calorimetry, is typically near −55°C for emulsion SBR and may shift upward in recycled rubber powders; this shift influences the low-temperature relaxation behavior of the blend. The overall compatibility criteria can be assessed by the fluorescence image uniformity, the EN 13399:2010 softening point difference, and the ratio of the complex moduli at low frequency after PAV aging.

In torch-applied and self-adhesive waterproofing membranes, the replacement of raw SBR emulsion with crosslinked SBR powder changes the viscoelastic response of the compound at application temperatures. Membrane compounds are typically produced in a horizontal kneader or a twin-screw extruder with L/D ratios of 32:1 to 48:1 at barrel temperatures of 160–190°C. The crosslinked powder does not melt in the extruder, so the screw must be configured with sufficient dispersive mixing elements; a screw with only conveying elements will produce poor powder distribution and uneven membrane thickness. The powder addition level for bituminous waterproofing compounds is commonly 5–12 wt% based on the total compound, depending on the filler loading and the desired cold flexibility. Cold flexibility measured by EN 1109:2013 is often used to evaluate membrane performance; the addition of coarse crosslinked powder can reduce cold flexibility more than fine powder at equal loading. Because the powder particles remain solid at application temperatures, the membrane surface may exhibit slight graininess, but this can be reduced by using powder with d50 below 120 µm and by adding a compatibilizing aromatic oil. Published data comparing specific crosslinked SBR powder grades in membrane formulations is limited, so pilot-scale trials are required before changing from emulsion to powder.