Why Nitrogen Knife Steels are Soft

Thanks to the new Knife Steel Nerds Patreon supporters! Hunter Chen, Claas, Curtis Fowler, Saptak Dutta, Tracker Dan, Paul Pedrazzi, 150LAT10N, James Kane, Vincent Franchino, and Jack Bashear.

I also have some Patreon-exclusive bonus content related to this article such as Nitrobe 77, Nitro-V, and 14C28N steel hardness data and experiments on the effects of the high temperature temper on nitrogen steel hardness.

Video

Shawn Houston and I discussed the results described in this video:

Nitrogen Steels

I have a previous article about what nitrogen steels are (there is no official definition) and how they are made so I’m not going to rehash that here. There are three main categories: 1) small nitrogen additions (<0.25% nitrogen) using relatively standard steelmaking methods, 2) Pressurized ESR production that can add higher amounts of nitrogen such as in LC200N with 0.40% nitrogen, and 3) high nitrogen powder metallurgy steels where the powder is nitrided before it is HIPed into a solid ingot. If you don’t know what those various terms mean then you can click on the link for the nitrogen steel article.

Carbon and Nitrogen and Steel Hardness

Carbon is the most common element used in steel to give hardness. The potential hardness of steel goes up with increasing carbon. The steel is heated up to high temperature, called austenitizing, and then is quenched to achieve the high hardness. The high temperature austenitizing puts the carbon atoms in an “interstitial” position, meaning between the iron atoms, and then rapidly quenching “locks in” those carbon atoms which distorts the room temperature structure of steel. That new distorted structure is called martensite. The more carbon is “in solution” in the steel the higher the hardness. In the chart below, for example, 0.4% carbon in solution equates to 60 Rc after quenching, 0.6% equals about 64-66 Rc, and 0.8% equals 65-67 Rc. Steel is typically tempered down to a lower hardness which also increases toughness. Carbon content is not the only factor that affects strength but it is a big one. You can read more about martensite and what gives it its high strength in this article.

Data from [1]

Nitrogen has a similar effect on steel as carbon. It is a similarly small atom and will be found interstitially in austenite and will lead to martensite formation during quenching. However, because it is a slightly heavier element, atomic mass of ~14 for nitrogen vs ~12 for carbon, more nitrogen is needed by weight to get the same level of strengthening. However, even after compensating for this, the maximum level of hardness achievable for a purely nitrogen-alloyed steel is a couple points less than for a carbon-alloyed steel.

Data adapted from [1][2][3]

Heat Treating Tests of Nitrogen Steels

I have performed a series of heat treating experiments with various nitrogen steels, and focused particularly on steels where I also have a steel with identical (or near identical) composition with no nitrogen intentionally added. This will allow better examples for what is happening in the steels when nitrogen is added to them. One example is BD1N steel, made by Carpenter and used by Spyderco, which was a nitrogen modification of their BD1 steel which was roughly a 440-type stainless steel.

According to Carpenter’s data there was an increase in the hardness of BD1N of about 2 Rc when compared with BD1 when given the same heat treatment. Below shows the hardness of each steel after austenitizing from different temperatures, quenching along with a cold treatment at -100°F, and then tempering at 300°F.

The BD1N steel behaves very similarly to the BD1 steel but the nitrogen addition means there is more nitrogen “in solution” after austenitizing which leads to higher hardness. Below shows ThermoCalc estimates of the carbon and nitrogen in solution from the different austenitizing temperatures used. The nitrogen is compensated for its higher atomic weight by multiplying by the ratio of atomic weight: 12/14.

Then if we plot the estimated C+N vs the measured hardness for the two steels we get a compelling correlation:

Adding Carbon Instead of Nitrogen

Why not just add more carbon to BD1 to get an increase in hardness? Below shows a hypothetical BD1 where the carbon has been increased by 0.12%, equivalent to the amount of nitrogen added for BD1N. The amount of carbon in solution is not increased as much when only carbon is added:

Another issue is that the “chromium in solution” is also dropped by adding more carbon to BD1. The chromium in solution controls the final corrosion resistance. So while in BD1N the chromium in solution was slightly increased relative to BD1, the carbon addition to the hypothetical BD1+C would lead to a decrease in corrosion resistance. This is similar to the trend seen in the 440 series of steels with 440A (0.7% C), 440B (.85% C), and 440C (1.05% C); the potential hardness goes up with carbon but the corrosion resistance also goes down.

The difference is that nitrogen is less prone to form nitrides with chromium than carbon is to form carbides. They are both hard particles that contribute to wear resistance but when more chromium carbides are formed that means less chromium is left in solution for corrosion resistance. During austenitizing those carbides/nitrides are dissolved to put the chromium in solution, so if there is more chromium carbide left in the steel after heat treating that means worse corrosion resistance. And some of that carbon goes toward forming of extra carbides which is partly why the carbon in solution doesn’t go up as much as nitrogen in solution would from a nitrogen addition. Nitrogen, while it can form chromium nitrides, has much less tendency to do so compared with carbon, so a 0.12% nitrogen addition to BD1 gives higher potential hardness and a small boost to corrosion resistance. So the nitrogen addition gives higher hardness without negatively affecting other properties.

Nitro-B vs 1.4116

Nitro-B steel, made by Buderus, is 1.4116/X50Cr15MoV but with a small nitrogen addition. It sees somewhat similar trends to adding nitrogen to BD1. Both are a small nitrogen addition to a pre-existing steel, where higher hardness levels are achieved vs the original. I austenitized both of these steels for 20 minutes followed by a plate quench, cryo in liquid nitrogen, and a 300°F temper. A 300°F temper is as low as I ever typically recommend for tempering so this is essentially maximum hardness for either steel.

You will notice that the 1.4116 steel tops out around 62 Rc and the hardness barely changes from about 2025 to 2150°F. This is a relatively low carbon steel at 0.5%, so once all of its carbide is dissolved, putting the full 0.5% carbon in solution, the hardness no longer increases. This hardness level is roughly in line with the carbon vs hardness chart shown earlier in the article (0.5% C = 62 Rc). Heat treating to 62 Rc is probably also impractical since with all of the carbide dissolved, grain growth would be a concern. Which is probably partly why heat treatment recommendations for 1.4116 typically top out around 1030°C (1885°F). So the nitrogen addition allows a higher level of hardness for Nitro-B because its combination of carbon and nitrogen in solution is greater. You will also notice that the temperature where 62+ Rc is reached is also lower because the C+N is put in solution at a lower temperature:

Carbon in solution

Carbon and (12/14)*Nitrogen

Nitro-B has slightly more carbon in solution for a given austenitizing temperature plus the 0.15% nitrogen, and as we see in the chart above the Nitro-B line is about 0.17% higher than the 1.4116 line. Then if we plot C+N vs hardness below we see that there is again a reasonably good correlation with hardness when comparing the predicted C+N with measured hardness:

You will also notice that the hardness of Nitro-B drops if austenitized above about 2000°F. The carbon and nitrogen in solution is increased when austenitizing higher so this is not the issue. Instead, the retained austenite becomes high enough that hardness is dropped. If you look at the carbon vs hardness chart from the beginning of this article you will also see that some of the datasets show a drop in hardness above a certain carbon level, about 0.9% carbon with those unalloyed steels. The temperature at which martensite forms during quenching is controlled by the carbon and alloy content. This includes a temperature where the martensite formation begins (martensite start) and a temperature where it is complete (martensite finish). When the finish temperature is below the quench temperature the transformation is incomplete, leaving some of the austenite phase from the austenitizing temperature; this remaining austenite is called “retained austenite.”

When retained austenite is greater than around 15-20% the hardness will be seen to drop as with Nitro-B. You can read more about retained austenite in this article about cryogenic processing. Using cryo means that more retained austenite is converted to martensite but you will reach a point where the hardness drops even when using liquid nitrogen. So the maximum austenitizing temperature when quenching to room temperautre is lower than when using cryo, and the maximum hardness is also increased when using cryo. For example, below is AEB-L showing the drop in hardness when quenching to room temperature vs cooling in a freezer or liquid nitrogen after the quench.

As-quenched hardness for Uddeholm AEB-L

The martensite start can be estimated with various models, such as this simple one below for the effects of carbon and chromium. The effect of nitrogen is also similar to carbon.

Ms (°C) = 539 – 423*C – 12.1*Cr

If we plot the predicted martensite start for AEB-L we see that when the Ms was dropped below about 350°F there was a drop in hardness even when using liquid nitrogen, and without cryo (room temperature quench), the hardness began to drop when Ms was below about 425°F:

So the 0.5% carbon and 15% chromium in 1.4116 does not lead to a reduction in hardness from excess retained austenite even at high austenitizing temperature, at least when using liquid nitrogen. There would likely be a peak in hardness and a drop when not using cryo. If we predict the Ms of 1.4116 we can also estimate where the steel would drop in hardness from excess retained austenite, assuming the peak hardness would be at a similar Ms to AEB-L (425°F). Based on those predictions, 1.4116 would likely have peaked in hardness around 1950-1975°F which would give it a peak hardness around 57-58 Rc. However, with Nitro-B you can get more carbon/nitrogen in solution, and those elements have the strongest effect on martensite start and finish temperatures and therefore even with the use of cryo you reach a point where the hardness goes no higher, and in fact will go down when using austenitizing temperatures above 2000°F

LC200N vs Nitro-B vs 1.4116

There is another nitrogen knife steel which is a modification of 1.4116 which we should discuss, however, which is LC200N. This steel was originally released under the name of Cronidur 30 though it is better known as LC200N these days, at least among knife enthusiasts. I have a full article about this steel, its history, and its properties here. LC200N is essentially 1.4116 but the 0.5% carbon is dropped to 0.3% and 0.4% nitrogen was added using PESR. The C+N is about the same as Nitro-B but the proportion of nitrogen is higher with LC200N.

When you plot out hardness vs austenitizing temperature for LC200N it matches Nitro-B at 1925°F but then drops with higher austenitizing temperature while Nitro-B continues to higher hardness up to 2000°F. This means that the peak hardness of LC200N is about 2 Rc lower than Nitro-B.

This is somewhat surprising as the composition of the two steels are basically the same apart from the higher proportion of nitrogen in LC200N. This reduces the austenitizing temperature range of LC200N but the peak hardness is actually found at a similar C+N for the two steels:

Here is the breakdown of the carbon and nitrogen in solution at different temperatures for reference:

Therefore the issue isn’t from more retained austenite forming in the LC200N than the Nitro-B since the peak hardness is found with roughly the same C+N. Because if it was from other alloying element effects (such as more chromium in solution), the peak hardness would have occurred at a lower C+N. So instead the most likely explanation is that the nitrogen is not increasing hardness as much as carbon, even after compensating for atomic weight, and therefore Nitro-B reaches higher hardness because of its higher proportion of carbon. With Nitro-B having similar chromium and molybdenum content there is a possibility that the corrosion resistance is also similar between the two steels. However, nitrogen can somewhat improve corrosion resistance on its own in stainless steels, so it is possible that LC200N would still have an edge in that area. This would be an interesting test to do in the future. If true, that would mean that Nitro-B could offer higher hardness but maintain a similarly high level of corrosion resistance which I would see as an upgrade. But another important comparison would be the microstructure of the two steels, as LC200N is quite fine while the 1.4116 I analyzed had some large carbides present in it, reducing toughness. This may have been from poor manufacturing of the 1.4116, however, as the largest carbides were particularly big. So it would also be useful to see how well Buderus is processing this steel and how fine Nitro-B’s carbides are.

LC200N microstructure is nice and fine

A giant carbide found in the 1.4116 steel I analyzed

Vanax vs Elmax

Uddeholm Vanax is very similar to their earlier PM stainless steel Elmax but with much of the carbon being replace with nitrogen:

However the heat treating response between the two steels is quite different and Elmax reaches significantly higher hardness levels:

Again there is a combination of factors at play here. One is that despite the similar chromium content (18%) the two steels have significantly different amounts of chromium in solution because of Vanax’s use of nitrogen. Nitrogen is much less prone to forming chromium nitrides than carbon is to forming chromium carbides, so the chromium is put in solution at a lower temperature. Below shows the chromium in solution for the two steels vs temperature according to ThermoCalc:

If you remember our simplified martensite start prediction which correlates with retained austenite, the chromium is a significant component:

Ms (°C) = 539 – 423*(C+N) – 12.1*Cr

Then we can plot the predicted martensite start for the two steels vs austenitizing temperature:

Vanax is reaching its peak hardness around a 1975°F austenitize which is a predicted martensite start of around 325°F, at which point the martensite start is low enough that excess retained austenite means a drop in hardness. This is roughly similar to the Ms we saw AEB-L start to drop in hardness. Elmax with its lower chromium in solution is still about 350°F Ms when austenitizing from 2150°F, which is probably near its peak hardness.

However, just like LC200N, the higher chromium in solution and retained austenite is not our only problem with achieving similar levels of hardness to Elmax, but also because the contribution of nitrogen to hardness is less than carbon. Below I have plotted the C+N for both steels vs hardness:

Curiously the expected hardness from a given carbon level is significantly higher than with Nitro-B/1.4116, however. For example, 0.4% carbon resulted in 63 Rc with Elmax but only 60 Rc with 1.4116. This could be from one or more of several factors, such as ThermoCalc not predicting the carbon in solution accurately, or the high carbide content in Elmax boosting hardness. If we assume that the ThermoCalc estimates are accurate, that would mean that Vanax would be capable of roughly 63 Rc if it used carbon alone instead of the combination of carbon and nitrogen.

Interestingly, the predicted carbon and nitrogen in solution is somewhat different between Vanax and LC200N despite the similar chromium in solution and both reaching similar peak hardness. Vanax reaches about 61 Rc at 1975°F in my testing where it had 0.2% nitrogen, 0.27% carbon, and 14.6% chromium in solution. LC200N reached peak hardness at 1925°F where it had 0.28% carbon, 0.33% nitrogen, and 14.7% chromium in solution. So this could be from differences in how well ThermoCalc is predicting the alloy in solution, or perhaps from differences in kinetics (the speed of the transformation). ThermoCalc is an estimate at “equilibrium” meaning an infinite hold time at temperature, while I am austenitizing for some relatively short period of time (15-20 minutes). And the amount of carbide/nitride that can be dissolved at 1925°F for LC200N would be less than would dissolve at 1975°F for Vanax. Also Vanax has vanadium nitrides along with its chromium carbides and the vanadium nitrides probably dissolve more slowly. In either case no model will ever be perfect so sometimes the calculations from ThermoCalc are more a guide than a true reflection of what would be measured experimentally.

Vancron vs CPM-10V

Uddeholm Vancron is very similar to CPM-10V except it has a significant amount of its carbon replaced with nitrogen. But when comparing the hardness of these steels they are actually quite similar:

So the nitrogen shifted the austenitizing range vs CPM-10V but the peak hardness is roughly the same. Why does Vancron reach very high hardness while the stainless nitrogen steels do not? The reasons are two-fold: 1) Vancron does not have high chromium like Vanax and LC200N. Those steels are designed to have high chromium in solution (~14-15%) for high corrosion resistance which increases their retained austenite. Vancron does not have this limitation. 2) Almost all of the nitrogen in Vancron goes towards the vanadium carbides/nitrides and not in solution for hardness. In fact ThermoCalc says that only about 0.03% nitrogen is in solution which is essentially zero. So the steel gets its hardness from carbon just like other non-nitrogen steels. Instead the nitrogen is used because vanadium carbonitrides have better resistance to adhesive wear than vanadium carbides. Unfortunately this is not an effect that is useful in knives which I wrote about in another recent article about Vancron.

Conclusions and Summary

So small nitrogen additions can be useful for improved hardness in grades such as BD1N or Nitro-B, or grades not discussed in this article like 14C28N. The nitrogen additions increase hardness by a similar mechanism to carbon but without negatively affecting other properties like corrosion resistance. But the high nitrogen, high corrosion resistance grades, LC200N and Vanax have limitations in hardness, with a max of about 61 Rc even when using cryogenic processing and very low tempering temperatures. This is from the combination of 1) high chromium in solution and therefore higher retained austenite, and 2) lower contribution of nitrogen to high hardness than carbon. With the recent Vancron article, and this one about the hardness potential of stainless nitrogen steels, it can seem like I am really being critical of nitrogen-alloyed steels. However, the element can be used in different situations in useful ways such as in the design of 14C28N or Nitro-B where I think it is very beneficial in those steel designs. It just depends on how the nitrogen is being used and for what, and of course the overall composition needs to be balanced with the nitrogen content.

---

[1] Krauss, George. “Martensitic transformation, structure and properties in hardenable steels.” Metallurgical Society AIME,(1978): 229-248.

[2] Tsuchiyama, Toshihiro, Kurato Inoue, Katsutoshi Hyodo, Daichi Akama, Nobuo Nakada, Setsuo Takaki, and Tamotsu Koyano. “Comparison of Microstructure and Hardness between High-carbon and High-nitrogen Martensites.” ISIJ International 59, no. 1 (2019): 161-168.

[3] Chiba, Makoto, Goro Miyamoto, and Tadashi Furuhara. “Microstructure of Pure Iron Treated by Nitriding and Quenching Process.” Journal of the Japan Institute of Metals76, no. 4 (2012): 256-264.