In boiling 90 percent nitric acid on a heating surface, the tests published by TIMET put unalloyed titanium at 1.0 to 6.6 mpy (0.03 to 0.17 mm/y) while Type 304L stainless ran 150 to 2,900 mpy (3.8 to 73.7 mm/y) in the same TIMET tests. That two-to-three order of magnitude gap is why nitric acid service is one of titanium’s strongest cases. But the same data show two things a designer has to know: the corrosion rate peaks in the middle of the concentration range rather than at the top, and red fuming nitric acid is a hard exclusion.
Titanium against stainless on a heating surface
| Metal temperature | Unalloyed titanium | Type 304L stainless |
|---|---|---|
| 240 F (116 C) | 1.1 to 6.6 mpy (0.03 to 0.17 mm/y) | 150 to 518 mpy (3.8 to 13.2 mm/y) |
| 275 F (135 C) | 1.6 to 6.1 mpy (0.04 to 0.15 mm/y) | 676 to 2,900 mpy (17.2 to 73.7 mm/y) |
| 310 F (154 C) | 1.0 to 2.3 mpy (0.03 to 0.06 mm/y) | 722 to 2,900 mpy (18.3 to 73.7 mm/y) |
Read the stainless column as the reason the equipment gets specified in titanium at all. Two honest limits on that comparison. Type 304L is not what a nitric plant would actually specify at this temperature: the real competitors are the silicon-bearing austenitics such as UNS S30600, and above roughly 80 percent acid, aluminium. And read the titanium column as what it is: a heating-surface condition, boiling, at one concentration, which is more severe than bulk liquid duty at the same temperature.
The counterintuitive part: the worst concentration is in the middle
Engineers new to titanium often assume that stronger acid means faster attack. In boiling pure nitric acid, liquid phase, the published curve does the opposite. Corrosion rate rises from near zero at 10 percent and falls again toward 70 percent, with the peak in between. On the TIMET curve the unalloyed and palladium-bearing grades crest near 24 to 26 mpy (0.61 to 0.66 mm/y) somewhere around 40 to 50 weight percent, while Grade 12 crests earlier and lower, near 15 mpy (0.38 mm/y), and stays lower right across the middle of the range.
The practical consequence: on that pure-acid boiling curve, a column or reboiler that spends its life at 35 percent acid is in a harder duty than one at 65 percent, and a process that swings through the middle band is harder than either. State the concentration range, not just the maximum.
The number to size the wall on
Before going further it is worth stating plainly what the published data do and do not agree on, because this page prints two figures for what looks like the same duty. For unalloyed titanium in boiling acid near 68 to 70 percent:
| Source within the same manual | Rate |
|---|---|
| The boiling-concentration curve for pure acid, read at 70 percent. Boiling liquid, unheated coupon; exposure duration not reported in the cited source | Roughly 5 mpy (0.13 mm/y) |
| The dissolved-titanium table, 68 percent boiling liquid with no titanium ion present, 24-hour test | 31.8 mpy (0.81 mm/y) |
These are different test references, and the spread is not an error in either. It is the real scatter in this service, and the manual’s own data elsewhere put boiling 70 percent acid across a range of roughly 2.5 to 37 mpy (0.064 to 0.94 mm/y), also without a stated exposure duration. The design rule that follows is the important part of this page: size the wall on the uninhibited pure-acid rate, on the order of 0.8 mm per year at the worst concentration, and treat metal-ion inhibition as margin rather than as design basis. The reason is in the next section: several places in a real plant see acid that carries no inhibiting ions at all.
Dissolved metal cations inhibit; other contaminants do the opposite
The effect is specific to dissolved metal cations. The manual names silicon, chromium, iron, titanium and precious metal ions such as platinum and ruthenium as inhibitors in minute amounts. It is not a general rule about contamination: chloride-bearing salts require their own data, while heavy nitrate salt loading moves the rate the other way and produces the worst straight-nitric numbers in the manual. Within that limit, in boiling 68 percent nitric acid small chromium contamination made Type 304L dramatically worse while titanium got better:
| Chromium in solution | Type 304L annealed | Unalloyed titanium |
|---|---|---|
| 0.0 percent | 12 to 18 mpy (0.30 to 0.46 mm/y) | 3.5 to 3.8 mpy (0.09 to 0.10 mm/y) |
| 0.0005 percent | 12 to 20 mpy (0.30 to 0.51 mm/y) | Not reported |
| 0.005 percent | 60 to 90 mpy (1.5 to 2.3 mm/y) | 0.9 to 1.6 mpy (0.022 to 0.041 mm/y) |
| 0.05 percent | 980 to 1,600 mpy (24.9 to 40.6 mm/y) | Not reported |
| 0.01 percent | Not reported | 0.1 to 1.4 mpy (0.003 to 0.036 mm/y) |
Dissolved titanium itself is an even stronger inhibitor of titanium corrosion in this service:
| Titanium ion added | In 40 percent HNO3 | In 68 percent HNO3 |
|---|---|---|
| 0 mg/l | 29.5 mpy (0.75 mm/y) | 31.8 mpy (0.81 mm/y) |
| 10 mg/l | Not reported | 0.8 mpy (0.02 mm/y) |
| 20 mg/l | 8.6 mpy (0.22 mm/y) | 2.4 mpy (0.06 mm/y) |
| 40 mg/l | 1.9 mpy (0.05 mm/y) | 0.4 mpy (0.01 mm/y) |
| 80 mg/l | 0.8 mpy (0.02 mm/y) | 0.4 mpy (0.01 mm/y) |
The manual applies this to recirculating streams such as stripper reboiler loops, where inhibition results from reaching a steady dissolved-titanium level. Two limits belong with that. First, reaching it is not automatic: it depends on wetted area, corrosion rate, loop volume and makeup rate, and the effect is self-limiting, because at 0.4 mpy the metal releases far less titanium than at 31.8 mpy. A high-makeup loop is closer to once-through than the word recirculating suggests. Second, and more important for safety, several parts of a real plant see acid with no inhibiting ions at all: once-through legs, startup, the period after an acid changeout or a wash, and condensate, which is by this same mechanism the purest acid in the system. Compared with an equilibrated recirculating loop, a laboratory coupon in fresh, uninhibited acid can read worse, not better. The once-through and transient locations just listed can resemble that uninhibited case, which is why the design basis above uses the uninhibited rate.
Vapour phase is a different duty from liquid
| Grade | Corrosion rate |
|---|---|
| Unalloyed titanium | 2.0 mpy (0.051 mm/y) |
| Grade 12 (Ti-0.3Mo-0.8Ni) | 0.8 mpy (0.020 mm/y) |
| Palladium-bearing unalloyed | 0.08 mpy (0.002 mm/y) |
Condensing acid and trickling condensate are the classic failure location in nitric service, and the grade ranking there is not the same as in the liquid. For hot, very pure acid and for condensing duty the published preference is the higher purity unalloyed Grade 1 (UNS R50250), and the reason is weldment behaviour rather than the base-metal rate in this table. Note also what this table is and is not: coupons suspended in vapour above the azeotrope for 144 hours. It is not a condensing-film test, and the manual attaches no rate to trickling condensate anywhere.
Where titanium is excluded outright
| Condition | Status |
|---|---|
| Red fuming nitric acid | Do not use. The published mechanism is not simple corrosion: attack is intergranular and leaves a surface residue of finely divided metallic titanium, which is highly pyrophoric and can detonate in the presence of a strong oxidising agent. The manual states the composition window for the reaction to develop: water content below 1.34 percent and NO2 above 6 percent, and it prints an explicit area of uncertainty between the reacting and non-reacting regions. Its appendix separately screens red fuming acid below about 2 percent water as ignition-sensitive; this exclusion therefore covers the uncertainty area out to that approximately 2 percent water line, not only compositions below 1.34 percent. Check your assay against that window rather than against the label, because a stream is rarely called red fuming on the datasheet. |
| White fuming nitric acid | Acceptable per the same manual, at roughly 0.1 mpy at room temperature. Do not read the red fuming exclusion as covering all fuming acid. |
| Nitric acid with chloride, up to aqua regia | Quantified hazard, and it reverses the usual grade logic. In the manual’s 96-hour boiling test of aqua regia, three parts hydrochloric to one part nitric, unalloyed titanium and the palladium-bearing grade each ran 44 mpy (1.12 mm/y), while Grade 12 ran 24 mpy (0.61 mm/y). The palladium-bearing grade therefore held no advantage over unalloyed, while Grade 12 was the lower of the three. At room temperature titanium is essentially immune to aqua regia. Rates in mixed acids rise as the reducing acid component or the temperature rises. |
| Nitric acid containing fluorides | Review required. Fluoride in acid can form hydrofluoric acid, which attacks titanium rapidly. |
| Very pure hot acid and vapour condensate | Allowed but grade-sensitive. Rates rise, weldments are the exposed feature, and higher purity unalloyed grades are the published preference. |
A supplier page that lists only the favourable rows is not usable for design. The red fuming exclusion in particular is a process safety item and belongs in the first conversation, not in a footnote.
What to put on the order
| Order line | What to state |
|---|---|
| Grade and standard | Grade 2 (UNS R50400) to ASTM B265 for plate or ASTM B338 for tube covers most nitric duty. Grade 12 (UNS R53400) where the mid-concentration boiling band governs. Non-condensing vapour: Grade 7 (UNS R52400). Hot very pure acid and condensing duty where weldment attack is the concern: Grade 1 (UNS R50250). Where both apply, Grade 11 (UNS R52250) is Grade 1 with the palladium addition |
| Service data | Concentration range in weight percent, not just the maximum; temperature; whether the surface is heated; liquid, vapour or condensing; recirculating or once-through |
| Contaminants | Fluorides, chlorides, and any metal ions already in the stream |
| Weld requirements | Filler grade and any weld examination, because weldments are the exposed feature in hot pure acid |
| Certificate | EN 10204 type 3.1 with heat and pour numbers |
One note from quoting this work: the question that changes the answer most often is not the acid strength, it is whether the wetted surface is heated. A heating surface at 90 percent boiling acid and a bulk liquid line at the same concentration are different duties and can end up as different grades.
Specification detail with sources
Corrosion values above are from the published titanium corrosion resistance manual of TIMET (Titanium Metals Corporation): corrosion of titanium and stainless steel heating surfaces in boiling 90 percent nitric acid at 240, 275 and 310 F; effect of chromium on corrosion in boiling 68 percent acid over three 48-hour periods; effect of dissolved titanium in 24-hour tests at 40 and 68 percent; resistance to nitric acid vapours over a 144-hour exposure above the boiling 70 percent azeotrope; the boiling-concentration curve for the three grades; the 96-hour boiling aqua-regia test; and the red fuming nitric acid caution. Grade identities are per ASTM: Grade 2 under UNS R50400, Grade 7 under UNS R52400 with palladium at 0.12 to 0.25 percent, and Grade 12 under UNS R53400 with molybdenum at 0.2 to 0.4 percent and nickel at 0.6 to 0.9 percent.
Scope note on how far these numbers carry: each row is a specific test, not a service prediction. The heating-surface rows are one concentration at three metal temperatures; the contamination rows are 48-hour periods with the acid replaced each period; the dissolved-titanium rows are 24 hours; the vapour rows are 144 hours above one azeotrope. Short tests in fresh acid tend to read worse than an equilibrated operating loop, which is exactly what the dissolved-titanium table demonstrates. Use these to rank grades and to know where the hard limits are, then confirm your own duty with the actual stream data. Beiyu Titanium supplies the plate, tube and bar to the standards named above with an EN 10204 type 3.1 certificate by order scope; the corrosion data is published third-party test data, not Beiyu test results.
Frequently asked questions
Which titanium grade for nitric acid service? Unalloyed Grade 2 handles most of the range. Grade 12 is the grade whose published boiling curve stays lowest where the duty sits in the mid-concentration band. Vapour and condensate are two different calls, and the source answers them with two different kinds of evidence. For non-condensing vapour the base-metal rate governs, and in the TIMET 144-hour vapour test the palladium-bearing Grade 7 (UNS R52400) ran 0.08 mpy against 0.8 for Grade 12 (UNS R53400) and 2.0 for unalloyed. For hot very pure acid and condensing duty the concern is not the base metal but accelerated attack at weldments, and there the published preference is the higher purity unalloyed Grade 1 (UNS R50250). Where a stream is both, Grade 11 (UNS R52250) is Grade 1 carrying the same palladium addition; Grades 16 and 17 are the leaner-palladium options.
Is stronger acid always worse? No, and this is the most common wrong assumption in this service. In boiling nitric the TIMET curve crests around 40 to 50 percent for the unalloyed and palladium-bearing grades and falls toward 70 percent, so a mid-strength stream can be harder duty than a strong one.
Can titanium be used in red fuming nitric acid? No. The published guidance rules it out because of pyrophoric reaction risk. Treat this as a process safety exclusion.
Why did my lab coupon read worse than the plant? Very likely the dissolved-titanium effect. In the TIMET 24-hour tests, going from 0 to 40 mg/l of titanium ion took the rate in 68 percent acid from 31.8 mpy to 0.4 mpy. Fresh acid in a short test has none of that inhibition.
What about reducing acids such as hydrochloric or sulphuric? That is the opposite case and needs a different grade. See Grade 7 titanium in reducing acids.
Which standard applies to the material? Plate and sheet run to ASTM B265, tube to ASTM B338, bar to ASTM B348. The plate guide is at the ASTM B265 sheet and plate guide.
Send the stream data with the RFQ
Send acid concentration range, temperature, whether the surface is heated, liquid or vapour phase, recirculation, contaminants, and the grade, standard, dimensions and quantity to the Beiyu RFQ page. Reviewed by Penn Ma, 15 years in titanium mill product export. Last reviewed: 2026-09-02.