You can read in text books that halide salts like sodium chloride and potassium chloride are corrosive towards HPLC systems. However, does anyone have any practical experience of this? Does the fact that HPLC mobile phases (hopefully) have little dissolved air and often use aqueous organic mixtures have any bearing on this use? I would not consider using a system routinely with halides since there are protein and ion chromatographs specially designed for these applications. Undoubtedly those working for protein or ion chromatography instrument companies will tell you that speciality systems are essential for such analyses. However, is there a sevice person out there who has actually seen a stainless steel HPLC system ruined by use of such salts? I'm referring to the use of RP-HPLC columns, not to low capacitity ion-exchange columns which perhpas are easily poisoned by metal impurities.

There are several previous threads on this topic which you might want to consult. In summary, the answer to your question is depends a lot on how often you plan on using halide containing eluents, what pH you are operating at, whether or not you're system is routinely maintained in terms of passivation and whether you are utilizing extremes of pH at different times on your instrument. Use of limited quantities of halides can be tolerated but this type of eluent system certainly requires more attention to passivation maintenance than is the case with other eluent systems. I can tell you that we make a DNA separation column which is frequently used with halide containing eluents and is frequently used on stainless steel instruments. While many of the users of this column report successful application on stainless steel instruments, column complaints are significantly higher when the columns are used on stainless steel instruments. Significant iron oxide contamination of the column is commonly observed in such cases.

But, I have a question for you: why not just use some other salt that isn't corrosive. There is certainly nothing magic about the chloride anion. For most purposes one can freely substitute other salts without any effect on the chromatography, even if the separation is via anion exchange. For example, methanesulfonic acid is available in high purity from Fluka. Its properties are nearly identical to the chloride except that it is noncorrosive and provides a much lower background in the ultraviolet.

Chris,

Your reply is extremely helpful and measured. Your suggestion about methanesulfonic acid also seems excellent. I would only note that in your reply you seem to have documented evidence that chloride ion can give rise to iron contamination of the (ion exchange) column. However, I have never seen hard evidence that chloride salts damage the instrument itself, at least in the pH range of say RP-HPLC (about 2.5-8). This would seem much less likely than poisoning of a delicate and expensive column.

Also a repeat: Check the material in question by inserting it in your chloride containing solution (it may be even feasable to check the outside of your column?). I have cut off sections of tubing used and seen absolutely no change of their surface in solutions of as much as 0.15M in NaCl (pH ~7!, PBS, etc.) for several month in a closed plastic test tube. I polish a section of the material and leave the rest as is. On the other hand, especially higher concentrations of HCl will immediatly turn stainless black (never have seen the red which, I think, was mentioned by Chris), so acidic Cl- seems to be an absolute NO- NO.

I realize that my earlier posting did not provide hard evidence of damage to the instrument itself but in one case I received a series of I think it was 10 or more different columns from the same customer who was interested in a diagnosis of the root cause of short column life. In this case, all of the columns were operated under alkaline conditions (I believe the pH was around 10) with sodium chloride in the 0.1-0.3 M range. Each of these columns contained considerable iron oxide (the red-brown color commonly seen on rusting steel components) with contamination extending throughout the entire length of the column. The magnitude of the iron oxide deposit in each column was considerable and I have no doubt that significant damage to the instrument occurred under these conditions. Furthermore, it is well known that metals will absorb to accessible silanols on reversed phase media, altering selectivity for species capable of interacting with metals. So, I don't think you can simply say that the problem is limited to ion exchange separations.

However, as noted above, if you stick to near neutral conditions and if you're system is well passivated, you should probably be able to use halide containing eluents with minimal damage to your instrument.

One additional point worth mentioning is the role of surface contamination in corrosion. A number of years ago I did the following experiment: I used stainless steel ferrules, some of which were used as is, some of which were passivated and the remainder of which were electropolished. I then placed all of three types of ferrules in 0.1M HCl. After a few hours the untreated ferrules had turned dark as mentioned above, even the passivated ferrules did the same thing although it took a day or two longer but the electropolished ferrules remain bright and shiny, even after months of exposure. The reason the electropolished ferrules could withstand these conditions while the other ones could not is that electropolishing removes tiny embedded particles of carbon steel and other foreign metals which contaminate stainless steel articles. These embedded particles arise from the manufacturing process of the stainless steel articles. Apparently only electropolishing can get rid of all of these minute particles. When such particles are present, they act as an electrode in an electrochemical cell, providing the electromotive power to induce corrosion of stainless steel under conditions which would not normally be observed. So, as you can see, when it comes to corrosion, life is not all that simple. Corrosion is indeed possible under conditions where stainless steel should be inert and since few instrument components are electropolished, one still has to worry about corrosion (especially piting corrosion) in the presence of halide containing eluents.

Of course itīs complicated, thatīs why I suggest to test the material. It just appears that there is an excellent chance that nothing happens. For many bilogical applications you would have to start from scratch (make tests that could take a lifetime) if you replace Cl-.

Chris, again: Was the brown stuff proven to be iron oxide? Did your customer introduce iron from the outside?

Hans,

Yes it was proven to be iron oxide. It was solubilized with oxalic acid as is the common practise in such cases but of course all that iron oxide had to come from somewhere.

But I guess I just don't see why you don't want to abandon corrosive eluents, even if they are weakly corrosive, why take the chance. As I mentioned before, I've faced that objection before and I can't remember a single case when a suitable substitute anion wasn't quickly identified.

I see "proven" a little different, but ok.
Like I said, I donīt take a chance, I test. Luckily nothing happened so, for example, it was possible to use PBS in SEC of antibodies, ie, under known conditions. Thatway I didnīt have to reinvent the world.

Thanks for all these answers-they are really helpful. The useful thing about sodium chloride is that you can add it to the mobile phase in a buffered eluent to alter the ionic strength without (hardly) altering the pH or the buffer capacity of the system. However, I guess sodium sulfate would do just as well without giving the corrosion problem.

Hans,

I guess I wasn't clear when I said it was proven in my previous post. I might have given you the impression that I ment it was proven simply by the fact that the contamination was removed by oxalic acid. It was proven to be iron oxide by analysis using ion chromatography and XPS. Iron oxides actually are found in a number of different forms with FeO and Fe3O4 being black but the more common form of iron oxide alpha-Fe2O3 is the more familiar red-brown color. Furthermore, the form generated under alkaline conditions with stainless steel equipment and chloride containing eluents is FeO(OH) which is brown in color. This is the form which we identified on columns used under alkaline conditions with NaCl eluents.

Chris, just read your 2nd post again (trying to resolve the discrepancy) and found that we almost appear to agree: "...near neutral conditions....should probably be able to use halide...."