Ferrosilicon Nitride vs Traditional Iron-Base Materials
Leave a message
Definition and Scope: What Is Actually Being Compared
Ferrosilicon nitride is a composite ferroalloy in which the functional phase is silicon nitride and the balance is free iron, residual unnitrided silicon and a small amount of oxygen. Commercial material is traded as FeSi3N4, and the grade is defined primarily by nitrogen content rather than by a single silicon figure, because nitrogen is the element that carries the effect the buyer is paying for. A refractory-grade lot typically sits at 28-32% N, 48-52% Si and 12-17% Fe, with oxygen held at 2.5% maximum and moisture at 0.3% maximum. A steelmaking-grade lot carries the same nitride phase but is supplied as a screened granule with a much tighter size band.
The phrase "traditional iron core materials" is used loosely in trade discussion for two different product families, and the comparison only becomes useful once they are separated. The first family is dense iron and low-carbon steel used for magnetic cores, counterweights and wear-facing bodies, where selection is driven by permeability, density, machinability and cost per kilogram. The second family is ferrosilicon and industrial silicon used as alloying additions, where selection is driven by silicon content, dissolution behaviour and recovery in the ladle. Ferrosilicon nitride competes with the first family on hardness and dimensional stability at temperature, and with the second on nitrogen delivery. A buyer who blends the two comparisons will usually specify the wrong grade.
This article compares the material on five axes that can be verified from a certificate of analysis and a size test report: composition control, hardness and hot stability, wear and corrosion behaviour, thermal conductivity, and cost structure. Each axis is treated separately because the ranking changes from one axis to the next, and that reversal is the single most important thing to understand before writing a purchase specification.
Chemical Composition and Commercial Grades
The specification that matters is not a single number but the ratio between the nitride phase and the metallic diluent. Nitrogen reports the nitride phase. Iron reports the metallic fraction, which is present because the material is made from ferrosilicon rather than from pure silicon, and which improves wettability and solution rate in liquid steel. Oxygen reports surface oxidation and residual silica. Moisture reports handling and storage quality. Two lots with identical nitrogen can behave differently at the point of use if their oxygen and moisture figures diverge, because both reduce the effective nitride fraction that actually enters the melt or the refractory matrix.
| Parameter | Refractory grade, typical | Steelmaking grade, typical | What it controls |
|---|---|---|---|
| Total nitrogen, N | 28-32% | 28-32% | Nitride phase fraction and alloying effect |
| Silicon, Si | 48-52% | 48-52% | Balance of nitride and free silicon |
| Iron, Fe | 12-17% | 12-17% | Density, wettability, solution rate |
| Oxygen, O | 2.5% max | 2.5% max | Surface oxidation, residual silica |
| Moisture, H2O | 0.3% max | 0.3% max | Storage quality, gas pick-up risk |
| Size | Powder, typically below 1 mm or 0-3 mm | Granule or lump, screened band | Packing density, dissolution, dust |
| Colour and form | Off-white to brown powder | Off-white granular | Indication of oxidation and nitriding degree |
Sampling discipline decides whether the certificate means anything. Ferroalloy lots are not homogeneous: nitrogen and oxygen segregate between fines and coarse fractions, and the fines concentrate oxidised material. Sampling should follow the general practice of GB/T 4010 for preparation of test samples, with increments taken from several positions in the lot and from the screened fractions in proportion to their mass. Nitrogen is normally determined by inert gas fusion after careful sample preparation, silicon by gravimetric or volumetric methods after dissolution, and oxygen by inert gas fusion on a separate aliquot. The acceptance, packing, marking and certification provisions of GB/T 3650 give the commercial framework: what the lot is declared to contain, how it is packed, and what document travels with it.
Where the material is used as a silicon source rather than a nitrogen source, it should be compared against the ferrosilicon grades of GB/T 2272 and the metallurgical silicon grades of GB/T 2881 rather than against iron. That comparison is on silicon content and cost per unit of silicon, and ferrosilicon nitride loses it immediately, because the nitriding step consumes energy and nitrogen without adding silicon. This is the first place where the ranking reverses depending on the axis chosen.
Hardness, Density and High-Temperature Stability
Hardness is where ferrosilicon nitride separates most clearly from iron and mild steel. Dense sintered silicon nitride is commonly quoted in the 1400-1600 HV range on a Vickers scale, and the indentations are measured on polished ceramic sections using the practice of ASTM C1327. The composite ferroalloy does not reach that figure, because free iron and residual silicon dilute the nitride network; its abrasion resistance therefore scales with the nitride fraction and with the degree of sintering, not with the headline nitrogen figure alone. Iron and low-carbon steel sit two orders of magnitude lower, near 120-180 HV in the annealed condition, which is why an iron body wears quickly in a sliding abrasive contact.
| Property | Ferrosilicon nitride, composite | Silicon nitride ceramic, dense | Iron / low-carbon steel |
|---|---|---|---|
| Hardness | Rises with nitride fraction and sintering degree | Approximately 1400-1600 HV | Approximately 120-180 HV, annealed |
| Density | Below the theoretical ceramic figure because of free iron | Approximately 3.2 g/cm3 | Approximately 7.85 g/cm3 |
| Melting behaviour | Nitride phase does not melt at ambient pressure | Dissociates rather than melts above roughly 1800-1900 C | Melts near 1500 C |
| Thermal expansion | Intermediate, dominated by the nitride phase | Approximately 3 x 10-6 per K | Approximately 12 x 10-6 per K |
| Electrical behaviour | Not intended as a magnetic or conductive path | Electrical insulator | Ferromagnetic and conductive |
The thermal expansion mismatch is the practical point. A nitride-bearing part bonded to a steel shell will see roughly a fourfold difference in expansion coefficient over the same temperature interval, and the resulting shear stress concentrates at the interface. Joints therefore need compliant interlayers, generous radii and controlled heating rates rather than a simple bolt-up. Iron core parts do not have this problem, which is one reason they remain the default where the duty cycle is thermal rather than abrasive.
High-temperature stability is often overstated in supplier literature. Silicon nitride does not melt at atmospheric pressure; it dissociates at roughly 1800-1900 C, and in air it oxidises well before that, with measurable scale formation generally reported from about 900-1000 C upward. The oxidation product is silica, which is protective up to a point but which consumes the nitride phase and, in the presence of slag or alkali, is fluxed away. A realistic service ceiling for an oxidising atmosphere is therefore set by the acceptable rate of phase loss, not by a single published temperature. Any figure quoted without the atmosphere, the holding time and the acceptable mass change is not a specification.
Wear, Oxidation and Corrosion Behaviour in Service
In sliding abrasion the nitride phase behaves as a dispersion of hard particles in a softer matrix. The hard particles shield the matrix until they are undercut, and then whole grains are pulled out, which is why the wear rate is sensitive to the strength of the bond between nitride and metal rather than to hardness alone. Fracture resistance should be assessed with the practice of ASTM C1421 and strength with the biaxial method of ASTM C1499 when the part is a structural ceramic; for a granular alloy these tests do not apply, and the relevant evidence is instead a service trial with a measured mass loss per unit of contact time.
In oxidation the ranking is more favourable than for iron. Iron oxides spall repeatedly because the oxide volume is roughly twice that of the metal consumed, so the scale ruptures and the metal keeps being exposed. Silicon nitride forms a vitreous silica layer that flows slowly and stays adherent, so the rate of attack falls with time as the layer thickens. The advantage disappears in the presence of fluorides, alkali oxides or heavy slag, all of which dissolve silica; in those environments an iron-base part protected by a refractory lining can outlast an unprotected nitride part.
In corrosion the advantage is real but narrower than trade articles suggest. Silicon nitride resists most mineral acids and is attacked by hydrofluoric acid and by hot concentrated alkali. Iron and steel corrode generally in humid and acidic service but can be protected by paint, galvanising or cathodic means at low cost, and that protection is often cheaper than the nitride premium. The honest comparison is therefore: nitride wins where the part is simultaneously abraded, hot and wet, and loses where only one of those conditions applies.
Thermal Conductivity and Cost: The Two Real Disadvantages
The thermal conductivity of dense silicon nitride is commonly quoted in the 15-30 W per metre-kelvin range at room temperature, against roughly 45-55 W per metre-kelvin for carbon steel. That is not an order-of-magnitude penalty, but it is enough to matter in a part that relies on conduction to shed heat, because the same heat flow needs a larger cross-section or a higher temperature difference. In a high-temperature wear application the lower conductivity is sometimes an advantage, since it keeps the backing structure cooler, and sometimes a defect, since it raises the surface temperature of the wearing face and accelerates oxidation. The direction of the effect has to be worked out per component rather than assumed.
Cost is the disadvantage that no property can offset. The cost structure of ferrosilicon nitride has four components: the ferrosilicon feedstock, which follows the silicon and electricity complex; the milling step that produces a powder fine enough to nitride in a reasonable time; the nitriding step itself, which holds the material at temperature in a nitrogen atmosphere for many hours and therefore consumes both energy and gas; and the screening and quality step that separates the size bands and certifies the nitrogen figure. Because furnace time is the bottleneck, unit cost is driven by throughput and by the number of heats per campaign rather than by raw material alone. This is why the material is priced per kilogram of contained nitrogen as much as per kilogram of product, and why a quotation that names only a tonnage price without the nitrogen specification cannot be compared with another.
Iron, by contrast, is produced in very large tonnage and its cost is dominated by ore, coke and energy at a scale ferrosilicon nitride will never approach. In any application where mass, section stiffness or magnetic permeability is the requirement, iron is cheaper by a wide margin and there is no technical case for substitution. The substitution case exists only where abrasive wear, hot oxidation or nitrogen delivery removes the cheaper material from service early.
Production Routes and Their Effect on Property Spread
The dominant industrial route is direct nitridation: ferrosilicon is milled to a controlled fineness, loaded into a batch furnace, and held under nitrogen at a temperature below the point at which the nitride phase would dissociate. Nitrogen diffuses into the powder and reacts with silicon to form silicon nitride, while the iron remains as a metallic network. Because the reaction is diffusion-limited, bed depth, gas flow, particle size and hold time all influence the nitrogen gradient across the batch, and a poorly controlled batch produces coarse particles that are under-nitrided at the core while the fines are fully converted. Screening does not fix that; it only hides it.
A second route is combustion synthesis, in which the exothermic nitridation of silicon is initiated locally and propagates through the compact without external heating. The route is fast and energy-efficient, but the product is a porous cake that must be crushed and graded, and the microstructure is coarser. A third route produces the nitride phase separately and blends it, which gives the best control of phase purity at the highest cost. For a buyer, the practical consequence is that two suppliers quoting the same nitrogen figure may be delivering different phase distributions, and a first order should always be qualified by a trial in the actual process rather than by certificate comparison alone.
For refractory use, the nitride is normally introduced into a taphole clay, a gunning mix or a silicon nitride bonded silicon carbide body, where it acts as an oxidation inhibitor and as a source of bond phase during sintering. Particle size and distribution control the packing density of the green body, and moisture control decides whether steam is generated during heat-up. For steelmaking use, the screened granule is added at tapping or in the ladle with argon stirring, where the nitride must dissolve rather than float, and where the free iron content is an asset rather than a diluent.
Selection Rules, Acceptance Testing and Storage
Four rules cover most decisions. Specify iron where the requirement is permeability, mass, stiffness or machinability, and accept the wear rate or protect the surface. Specify ferrosilicon where the requirement is silicon addition and nitrogen is irrelevant, and use the grades of GB/T 2272. Specify ferrosilicon nitride where nitrogen must be delivered together with silicon, as in nitride-strengthened rebar produced to GB/T 1499.2 or in grain-oriented silicon steel produced to GB/T 2521.1, where nitride and aluminium nitride type inhibitors control grain growth. Specify a dense silicon nitride ceramic, not the ferroalloy, where the part is a structural component that must survive thermal cycling without a metallic binder.
Acceptance should be built on a certificate that states nitrogen, silicon, iron, oxygen and moisture, the size band with its tolerance, and the sampling basis. The size band should be verified by screening a representative sample rather than by visual inspection, because fines are the fraction that oxidises fastest and the fraction most likely to be under-reported. A moisture determination on the as-received sample is cheap and catches wet material before it reaches a furnace.
Storage is part of the specification. The material should be kept in sealed packing, off a concrete floor, under cover, and rotated on a first-in first-out basis. Opened bags should be closed or consumed within the shift, because nitride fines pick up moisture and oxidise at the surface, which raises the oxygen figure and lowers the effective nitrogen at the point of use. Stacked bags should not be compressed to the point of rupturing the liner, and the store should be separated from acid and alkali sources. A lot that has been open for a season should be re-analysed before use rather than assumed unchanged.
FAQ
Q: What nitrogen content should a purchase specification state for ferrosilicon nitride?
Commercial refractory and steelmaking grades are typically specified at 28-32% total nitrogen with silicon at 48-52% and iron at 12-17%. The nitrogen figure should always be paired with an oxygen limit, usually 2.5% maximum, and a moisture limit, usually 0.3% maximum, because both reduce the effective nitride fraction.
Q: Is ferrosilicon nitride harder than iron?
Yes, substantially. Dense silicon nitride is commonly quoted at roughly 1400-1600 HV, while annealed iron and low-carbon steel sit near 120-180 HV. The composite ferroalloy falls between those figures depending on its nitride fraction and sintering degree, so a hardness claim should be tied to a measured sample rather than to the generic material name.
Q: Why is ferrosilicon nitride more expensive than iron or ferrosilicon?
Because the nitriding step holds milled ferrosilicon at temperature under a nitrogen atmosphere for many hours, so furnace time, energy and gas dominate the cost rather than raw material. Unit cost therefore depends on throughput and campaign length, which is why quotations should be compared per kilogram of contained nitrogen.
Q: Does the low thermal conductivity of silicon nitride always cause problems?
No. Dense silicon nitride is commonly quoted at 15-30 W per metre-kelvin against 45-55 for carbon steel, so a part that must conduct heat away needs more section. In a wear application the same low conductivity can protect the backing structure by keeping it cooler, but it also raises the surface temperature of the wearing face and can accelerate oxidation.
Q: At what temperature does ferrosilicon nitride fail in air?
There is no single failure temperature. Silicon nitride dissociates rather than melts at roughly 1800-1900 C at atmospheric pressure, but in air it oxidises well before that, with measurable scale formation generally reported from about 900-1000 C. The usable ceiling depends on the acceptable rate of nitride loss over the required service life.
Q: How should a delivered lot be checked before it enters the process?
Check the certificate against the purchase specification, screen a representative sample to confirm the size band and the fines fraction, and determine moisture on the as-received sample. Sample preparation should follow the general practice of GB/T 4010, and packing, marking and documentation should follow GB/T 3650. A first lot should also be qualified by a trial in the actual furnace or ladle.

