How Tungsten Carbide Powder is Made: From Ore to Powder

This flowchart shows two parallel paths to tungsten carbide powder. The virgin route (left side) starts from ore, concentrates it, converts to APT (ammonium paratungstate), then through oxide and metal stages to WC powder. The recycling route (right side) uses the zinc process to reclaim WC-Co directly from scrap, bypassing the ore-to-APT steps entirely. Both routes can feed into RTP (ready-to-press) production. The zinc process recovers 95%+ of tungsten at about one-third the energy cost of virgin production, which is why recycling provides ~30% of global tungsten supply. Understanding these routes helps explain price differences: recycled material is cheaper but may have trace impurity variations.

Each manufacturing step adds cost and affects final powder properties. Notice the temperature progression: starting cold for mechanical concentration, then escalating through chemical processing (80-100°C for APT), thermal decomposition (500-800°C for oxide), hydrogen reduction (700-1000°C for metal), and finally carburization at the highest temperatures (1300-1700°C for WC). The reduction and carburization temperatures directly control grain size—lower temperatures within each range produce finer powders because grain growth is thermally activated. This is why submicron WC powders require precise temperature control and cost 2-3x more than coarse grades: the process windows are tighter and yields are lower.

Tungsten occurs primarily in two mineral families. Wolframite is an iron-manganese tungstate that's dark and dense, easily separated by gravity methods, and processed through alkali digestion (NaOH) that dissolves the tungsten while leaving iron behind. Scheelite is calcium tungstate with higher theoretical WO₃ content, but it's harder to separate from other calcium minerals and requires either acid or alkali processing depending on ore composition. Secondary ores at 0.1-1% WO₃ are increasingly important as high-grade deposits deplete—they require extensive concentration before chemical processing, adding significant cost. The choice of processing route (acid vs. alkali) depends on which impurities need removal and affects downstream APT quality.

The extreme concentration of tungsten production in China—over 80% of global mine output—creates significant supply chain risk. China also controls over 85% of global refining capacity, meaning even ore mined elsewhere often goes to China for processing. Recent export controls (effective February 2025) require special permits for tungsten exports, causing price spikes and supply uncertainty. This concentration drives two strategic responses: investment in recycling (now ~30% of consumption) and development of non-Chinese sources in Vietnam, Portugal, and Austria. For purchasers, this means understanding your supply chain: is your powder from virgin Chinese material, recycled scrap, or alternative sources? Each has different risk profiles and price dynamics.

Ore concentration is a multi-stage process that exploits physical property differences. Crushing and grinding liberate tungsten mineral particles from the surrounding rock. Gravity separation takes advantage of tungsten's high density (wolframite: 7.1-7.5 g/cm³, scheelite: 5.9-6.1 g/cm³) to separate it from lighter gangue minerals. Flotation uses chemical reagents to selectively attach bubbles to tungsten minerals, floating them to the surface. Magnetic separation is particularly effective for wolframite because it's paramagnetic while most gangue is not. The progressive grade improvement (from 0.1-1% ore to 65-75% concentrate) comes at a cost: overall tungsten recovery is only 40-60%, meaning significant tungsten is lost to tailings at each stage.

Alkali digestion is the traditional route for wolframite ores. The key chemistry involves dissolving tungsten into sodium hydroxide solution while leaving iron and manganese behind as insoluble oxides. The sequence matters: first dissolve the tungsten (digestion), then remove impurities through a series of precipitation and filtration steps, then convert sodium tungstate to tungstic acid by adding HCl, and finally neutralize with ammonia to crystallize APT. Each purification step targets specific contaminants—molybdenum is particularly problematic because it follows tungsten through most processing steps and must be removed by selective precipitation. The multi-step process explains why APT purity grades exist: more purification steps = higher purity = higher cost.

Acid digestion is preferred for scheelite ores because calcium dissolves readily in hydrochloric acid. The challenge is different impurities: scheelite ores often contain phosphorus and arsenic that must be removed through solvent extraction (using organic solvents that selectively pull impurities from the aqueous tungsten solution). Ion exchange columns then convert the purified tungstate to ammonium form for crystallization. This route generally produces higher-purity APT than alkali digestion because solvent extraction is more effective at removing trace contaminants. Modern APT plants often use both routes, selecting based on concentrate composition and purity requirements.

APT grades reflect the downstream application requirements. Technical grade (99.9% purity) is adequate for most carbide production where trace impurities below 50-100 ppm don't affect sintering or final properties. High-purity grade (99.95%) is required for fine-grain carbides where even small impurity levels can act as grain growth sites or create defects at the microscale. Ultra-high purity (99.99%) is reserved for specialty applications like electronics-grade tungsten where parts-per-million contamination causes functional failures. Each purity jump roughly doubles the price because additional purification steps have diminishing yields. When specifying powder, consider whether you actually need higher APT purity—paying for 99.99% when 99.9% works is wasted money.

The oxide stage is a critical control point for final particle size. Yellow oxide (WO₃) is the fully oxidized form, stable and easy to produce in air—but it produces relatively coarse tungsten upon reduction. Blue oxide (WO₂.₇₂) is partially reduced and has a unique needle-like crystalline morphology; during hydrogen reduction, these needles fragment into extremely fine tungsten nuclei, making blue oxide the preferred starting point for submicron WC production. Brown oxide (WO₂) is an intermediate that's only stable under reducing conditions. The color difference is actually useful: operators can visually verify they're getting the right oxide phase. Choosing yellow vs. blue oxide is one of the first decisions that determines final WC grain size.

This table reveals why fine powder production is so challenging. Every parameter must work together: low temperature slows atomic diffusion and grain growth; high hydrogen flow sweeps away water vapor before it can drive the CVT (chemical vapor transport) mechanism; shallow bed depth prevents water accumulation; short dwell time minimizes growth opportunity. The CVT mechanism is key: water vapor reacts with WO₂ to form volatile WO₂(OH)₂, which deposits on existing tungsten particles, causing them to grow. Controlling water vapor is actually more important than temperature for preventing coarsening. This is why rotary kilns (which constantly tumble the powder, preventing H₂O accumulation) can produce finer powder than static pusher furnaces at the same temperature.

Furnace selection involves trade-offs between control precision, throughput, and capital cost. Pusher furnaces offer excellent temperature control (±5°C) and can produce the finest powders (0.5 μm), but throughput is limited. Rotary kilns sacrifice some precision for dramatically higher throughput (500-2000 kg/day); the continuous tumbling action prevents water vapor accumulation, which partially compensates for looser temperature control. Fluidized beds offer the best of both worlds—excellent temperature uniformity and good water vapor removal—but are complex to operate and limited to certain particle size ranges. Batch furnaces are flexible but low-throughput, mainly used for specialty grades. Most large producers use rotary kilns for standard grades and pusher furnaces for premium fine grades.

Note the inverse relationship between particle size and oxygen content: finer powders have more surface area and therefore more oxygen pickup. Ultra-fine tungsten (0.3-0.5 μm FSSS) typically contains up to 1500 ppm oxygen—this oxygen must be accounted for in carbon balance calculations during carburization. The carbon specification (<100 ppm) is tight because residual carbon from oxide reduction must not interfere with subsequent carburization stoichiometry. Iron limits increase with coarser grades because the impact of contamination is proportionally lower on larger particles. When specifying tungsten powder, the oxygen level is as important as particle size: high-oxygen powder requires more carbon during carburization to compensate.

Carburization temperature must match the tungsten particle size to prevent unwanted grain growth. Fine tungsten powder (<0.5 μm) carburizes at the lowest temperatures (1300-1380°C) because the high surface area provides adequate reaction kinetics even at lower thermal activation. Coarse powders (>4 μm) require temperatures up to 1700°C because carbon must diffuse further into larger particles. The critical constraint: if carburization temperature is too high for fine powder, grains will grow and you'll lose the fine structure you paid premium prices to create. This is a classic optimization problem: temperature must be high enough for complete W→WC conversion but low enough to prevent coarsening. Carbon black is preferred for fine grades because its small particle size ensures intimate contact; graphite works for coarse grades where mixing is easier.

Stoichiometric WC contains exactly 6.13 wt% carbon (one carbon atom per tungsten atom). Achieving this target requires accounting for oxygen in the tungsten powder (which consumes carbon as CO during heating), carbon from the reaction atmosphere, and analytical precision in both carbon addition and measurement. Carbon deficiency leaves unreacted W₂C (ditungsten carbide)—a different phase with inferior properties that causes sintering problems. Carbon excess leaves free graphite that creates porosity and reduces strength. The acceptable window is typically 6.10-6.20% total carbon with <0.05% free carbon. XRD can detect W₂C phases; magnetic property measurements (tungsten and W₂C are magnetic, WC is not) provide fast quality checks.

Carburized WC comes out of the furnace as sintered lumps—not individual particles. Deagglomeration breaks these lumps back into individual grains without damaging the grain structure. Jaw crushing handles the initial size reduction from furnace-sized lumps to millimeter-scale pieces. Ball milling breaks these further while simultaneously narrowing the particle size distribution through preferential breakage of oversized particles. Jet milling uses high-velocity gas to cause particle-particle collisions, effective for ultrafine grades but expensive to operate. Classification (air separation or screening) removes out-of-spec material—oversized particles that would create defects, or ultrafines that would interfere with pressing. The milling process can introduce iron contamination from grinding media; ceramic-lined mills minimize this but add cost.

This suite of tests characterizes powder quality comprehensively. FSSS particle size is the primary specification that customers order against. Total carbon and free carbon together verify proper carburization—total C should be 6.10-6.20% with minimal free carbon. Oxygen indicates surface oxidation, which affects sintering behavior and carbon balance during downstream processing. Iron contamination typically comes from milling equipment and must be controlled to avoid forming iron-cobalt intermetallics during sintering. BET surface area provides a second check on particle size (finer particles have higher surface area) and is especially important for ultrafine grades where FSSS measurements have higher uncertainty. A complete certificate of analysis (COA) includes all these parameters.

The zinc process is elegantly simple and highly efficient. Molten zinc at 850-950°C dissolves into the cobalt binder phase, forming a zinc-cobalt alloy that swells dramatically (200% volume increase). This expansion breaks apart the rigid WC skeleton without chemically altering the WC particles. At higher temperature (1000-1050°C) and vacuum (6-13 Pa), the zinc evaporates and is recovered for reuse—zinc recovery exceeds 99%. What remains is a friable "cake" of WC and cobalt that crumbles easily and is ball-milled to powder. The WC particles emerge essentially unchanged from their original structure, retaining their grain size and morphology. This is why zinc-reclaimed powder can be directly reused for the same grade as the original scrap, with only minor composition adjustments.

Chemical recycling takes a different approach: instead of preserving the WC structure, it converts scrap entirely back to APT—the same intermediate used in virgin production. Oxidation roasting at 800-1000°C converts WC to WO₃ and cobalt metal to CoO. Alkali leaching dissolves the tungsten oxide while leaving cobalt as a recoverable residue. The tungstate solution goes through the same purification steps as virgin APT production. This route has lower tungsten recovery (85-90% vs 95%+ for zinc process) and higher energy consumption, but produces virgin-equivalent material that can be used for any grade. Chemical recycling is preferred when scrap contains multiple carbide types (WC-TiC-TaC) or when a complete composition change is needed.

This economic comparison explains why recycling has grown to 30% of tungsten supply. Virgin production from ore requires 12-15 kWh per kg of tungsten, with only 60-70% recovery from ore to final product due to losses at each processing stage. Zinc reclamation uses just 3-5 kWh/kg—a 70%+ energy reduction—with 95%+ recovery. The output quality is excellent for remaking the same grade, though contaminant accumulation limits how many times material can be recycled through the zinc route. Chemical recycling at 8-10 kWh/kg costs more than zinc but less than virgin, with the advantage of producing virgin-equivalent material that can be used for any grade. Supply security, not just cost, drives recycling investment: recycled tungsten isn't subject to Chinese export controls.

Understanding cost breakdown helps explain price differences between suppliers and grades. Mining and concentration (15-25%) varies enormously with ore grade—a 1% ore costs twice as much to process as a 2% ore. Hydrogen reduction (15-20%) is energy-intensive and sensitive to H₂ prices, which vary regionally. Carburization (15-20%) consumes significant energy at 1300-1700°C furnace temperatures. For fine powders, reduction and carburization costs are higher because temperature control is tighter and yields are lower—explaining the 2-3x premium for submicron grades. QC testing (10-15%) is a larger percentage for premium grades requiring comprehensive analysis. Overheads (15-25%) include the substantial working capital tied up in inventory—tungsten's high value means inventory costs matter.