Sometimes learning more about a common heat-treat phenomenon can yield a deeper understanding of the underlying principles involved. Such is the case when The Doctor was asked a simple question: What is the difference between the formation of surface decarburization and intergranular oxidation in atmosphere-furnace processing? Let’s learn more.


The Answer

The basic difference is that decarburization – partial or total – is the result of oxygen combining with carbon at the surface and the associated diffusion of carbon to the surface, such that a gradient of carbon from the surface to some interior depth occurs. The result is a layer largely devoid of carbon that transforms (primarily) to ferrite on quenching.

By contrast, intergranular oxidation/intergranular attack (aka IGO/IGA or grain-boundary oxidation) occurs primarily along grain boundaries (areas of low lattice energy) due to elements that oxidize preferentially to iron, namely silicon, chromium and manganese. This results in the loss of these elements in the near-surface region. This lowers the hardenability of the affected zone and results in intermediate transformation products (e.g., bainite, pearlite) in the localized microstructure, significantly affecting the fatigue strength in these local zones. It is especially important in gears, bearings and other carburized mechanical-systems components. 


Digging Deeper

Decarburization is a well-known kinetic-driven process involving the removal of near-surface carbon from steel at elevated temperatures primarily (but not exclusively) within the austenite phase region. Contributing factors are high temperature and large amounts of oxygen present in the furnace atmosphere. Under these conditions, carbon atoms migrate to the surface and are oxidized. CO or CO2 is formed and escape as gases.

As a result of near-surface carbon loss, the surface of the steel upon transformation has lower tensile strength and hardness. The fatigue resistance, wear rate, residual-stress state and case hardness are also adversely affected.[1,2]

Decarburization can be caused by the absence or loss of protective atmosphere, a furnace atmosphere whose carbon potential is less than the surface carbon content of the parts or by air infiltration into the furnace. It can be minimized by proper atmosphere control, having a “tight” furnace and surface protection (e.g., copper plating). Decarburization occurs to some extent whenever a part is run in an endothermic-gas atmosphere.

Complete decarburization means an area with almost total removal of carbon. Partial decarburization (Fig. 1) is to be understood as the area in which a reduction of the carbon content has occurred but which has not resulted in total removal of carbon. During microscopic evaluation, the decarburization depth is determined on the basis of a change in microstructure.[3]

Decarburization is a persistent problem during high-temperature heat treatments carried out in industrial operations, such as forging and rolling operations. As a result, machining operations are often necessary following heat treatment to remove the near-surface layer of the material.


Calculating the Decarburized Layer[1]

The hardness of tempered martensite is considered a linear function of the carbon content for a given iron-carbon alloy system in the range of 0.2-0.6% C. Decarburization involves surface reactions and carbon diffusion, so the (formal) diffusion equation provides a practical and accurate mathematical model. Decarburization obeys Fick’s second law of diffusion (Eq. 1). It is well known that the diffusion coefficient follows an Arrhenius relationship and is exponentially dependent on temperature. Different alloying compositions can also greatly affect the decarburization response of a particular type of steel.


(1)     (Cx – CO)/(Cs-CO) = 1 - Erf(x/(2(Dt)1/2)


Cx is the carbon concentration at a distance x from the surface

CO is the carbon concentration at the core

CS is the carbon concentration at the surface

Erf is the error function

D is the diffusivity of carbon in the alloy

t is the diffusion time


A Fourier series (Eq. 2) can be used to predict decarburization. This series can be evaluated using advanced software such as Mathematica® to determine the depth of decarburization. It has been found that 15 to 20 terms of the Fourier series are necessary to obtain the desired accuracy of results.[1]




Taking Corrective Action

Decarburization can be prevented during heat treatment by copper plating, wrapping a component in a protective stainless steel foil or by use of paints specifically designed for this purpose. In addition, carbon restoration cycles (Fig. 2) are highly effective and can be performed in an atmosphere furnace by raising the carbon potential to restore the near-surface carbon layer.


A Few Words about IGO/IGA

Intergranular oxidation (IGO) and intergranular attack (IGA) are often considered similar phenomenon to decarburization, but they really are not. IGO/IGA is a very shallow surface-layer phenomenon (Fig. 3) typically observed in the range of less than 0.13 mm (0.0005 inch), which can negatively affect mechanical properties (e.g., bending fatigue life). In rare instances, layers as deep as 0.00075 inch or more have been observed.

IGO/IGA is caused by oxygen present during atmosphere hardening or carburizing. The type of oxide formation will depend on the alloy content of the steel. At the grain boundaries, one finds oxides of chromium, molybdenum, vanadium and other elements. The degree to which these oxides form will depend on the process temperature and the time at temperature. IGO/IGA can be minimized by good atmosphere control, having a “tight” furnace (i.e., the absence of air leaks) or the use of low-pressure (vacuum) carburizing. Grinding of the layer is typically performed to remove it.

During hardening or carburizing, oxygen atoms, which are approximately 35% smaller than iron atoms, are present as a direct result of endothermic-gas decomposition. The oxygen diffuses slowly into the steel surface (due to the solubility of oxygen in iron), migrating to and along the grain boundaries. The rate of diffusion is dependent on the oxygen potential of the furnace atmosphere and the process temperature. The origin and source of the oxygen comes from the endothermic process gas plus enrichment gas plus dilution air, if used.



Even seemingly simple heat-treat phenomena need to be understood in greater detail to provide better understanding of the principles, methods and mechanism involved, and the ways in which to mitigate them should the need arise.



  1. Cioffi, Rebecca D., “A Comparison Study on Depth of Decarburization and the Role of Stable Carbide Forming Elements in 1075 Plain Carbon Steel and 440A Stainless Steel,” Rensselaer Polytechnic Institute, Department of Materials Science and Engineering
  2. Herring, Daniel H., “Atmosphere Gas Carburizing – Case Studies, Lesson’s Learned,” Industrial Heating (Part 1), Sept. 2012, (Part 2), Oct. 2012
  3. B. Edenhofer and K. Voges, Section Decarburization, “Carburizing of Steels,” Thermochemical Surface Engineering of Steels, 2015
  4. Baggerly, R. G. and R. A. Drollinger, “Determinaton of Decarburization in  Steel,” Journal of Materials Engineering and Performance, Volume 2, Issue 1 Feb. 1993, p. 47-50
  5. SAE Recommended Practice, “Methods of Measuring Decarburization,” SAE J419, 1990 SAE Handbook, Materials, Vol 1, Society of Automotive Engineers, 1990, p 3.12-3.16
  6. H. Burrier, Jr., “Hardenability of Carbon and Low-Alloy Steels,” Metals Handbook, Vol 1, 10th ed., ASM International, 1990, p 464-484.
  7. R.A. Grange, C.R. Hribal, and L.F. Porter, “Hardness of Tempered Martensite in Carbon and Low-Alloy Steels,” Metall. Trans. A, Vol 8, 1977, p. 1775-1785
  8. J.H. Brophy, R.M. Rose, and J. Wulff, The Structure and Properties of Materials, Volume II: Thermodynamics of Structure, John Wiley & Sons, 1964, p 75-97
  9. “Hardenability Curves,” Metals Handbook, Vol 1, 10th ed., ASM International, 1990, p 535-543
  10. Herring, Daniel H., “Decarburization,” Industrial Heating, May 2016
  11. Chatterjee-Fischer, Ruth, Internal Oxidation During Carburizing and Heat Treating, Met Trans A , Vol 9A, ASM International, 1978, pp. 1553-1560
  12. Dowling, W.E. et al, Bending Fatigue Behavior of Carburized Gear Steels : Four Point Bend Test Development and Evaluation, SAE Technical Paper 960977, 1996