Cleaning AFM Tips With Hydrofluoric Acid and Safety Comments on Working With HF

I recently cleaned a bunch of old Digital Instruments Si TESP tapping mode AFM probes by submerging them in buffered HF for a few seconds. Although the process is tricky, most came out sharper than they were before the process. (HF is extremely dangerous; don't try this without completely educating yourself on the hazards of HF, which are discussed to some extent below.) Etching an entire wafer of tips will not be recommended by any manufacturer because it is unlikely to help the tip aspect ratio. Acidic fluoride etches are not face-specific and though they etch unoxidized Si very slowly, they do so indiscriminately. Thus, etching a Si probe with no oxide on it would only serve to reduce the tip aspect ratio and increase the effective tip diameter. (There are face-specific etches; but I do not know if, for example, face-specific aqueous caustic etches have been used to re-sharpen tips. Might be worth trying.) The reader may now be asking why I would suggest doing anything to their tip that may endanger its precious aspect ratio! To make sense of this suggestion, following key facts are needed:

1) Aqueous acidic fluoride (HF) solutions etch SiO2 much more rapidly than they do unoxidized Si, by several orders of magnitude.

2) Si oxidizes slowly but surely upon exposure to oxygen-containing air. Faster if that air is wet, and much faster if the air is wet and the Si is exposed to light more energetic than the Si bandgap (UV and visible radiation, actually any electromagnetic radiation with a wavelength shorter than 1100 nm).

3) The Si atoms in SiO2 are appreciably less densely packed than they are in Si, so as an oxide layer grows, the oxide expands to fill a volume larger than that of the Si from which it grew.

Put these three facts together and maybe I don't seem quite so loony. If one has Si tips that have been sitting around for a good while, they will have grown an oxide surface layer. Their diameter will have grown slightly, at the expense of both the effective tip radius and the aspect ratio. Additionally, things may have grown or settled or creeped onto the surface of the tips, onto the surface of said oxide. I certainly would try using any given tip before carrying out the following process. But if a given tip is terrible one can't go wrong by giving this a shot. It is more likely to work the older the tip is. It is not likely to be very helpful more than once or twice, though rinsing tips in solvents using a similar process sometimes is.

1) Completely submerse the Si probe in a solution of buffered HF or diluted HF for about 30 seconds, making sure to get all air bubbles off of it. This is most readily done with polymer tweezers (Delrin or equivalent), with notches cut into the tip to hold the tip tightly by its sides. (See Figure) Without notches it is all too easy to have the tip flip onto its flat, destroying it.

Try to put the probe in tip-first, at about a 45 degree angle. Wave the tip around in the etchant solution to any get bubbles off, but don't let go of it, it's very annoying to pick up again.

2) Take the tip out of the solution and rinse it with copious amounts of the purest water available, at least deionized. Rinse into a plastic bottle, which now contains hazardous HF waste. Very dilute HF is allowed in most sewer systems; after all, most water is fluoridated. However, high fluoride concentrations in sewer effluents is a common and serious problem and if one dumps HF waste with abandon the local water treatment folks will pay an unpleasant and well-deserved penalty assessment visit.

3) Use compressed air or nitrogen at high pressure (I use a 100psi N2 gun) to thoroughly dry the tip, still holding it tightly with the tweezers. It is very important to dry the area between the tip and the bulk of the probe, so it doesn't cement the tip to the probe such that it no longer resonates.

4) If working in a tapping/oscillating tip mode, the tip may have to be driven with a large oscillation piezo voltage to free it, if the tip isn't rinsed with very pure water and dried with a high-velocity gas stream. Set the drive amplitude to ~5 V, then autotune. Buffered HF is available from semiconductor specialty chemical houses. Concentrated HF(aq) is readily available and "diluted HF" is just concentrated HF diluted 1:10 with distilled (or better, Nanopure) water.

*Extremely important safety comments on the hazards of hydrofluoric acid:*
This is just what I have to say. HF is dangerous so don't just go on what some sap like me has to say. Educate *yourself* before you use HF!

HF is a weak acid. It is not its acidity per se that presents a risk to those exposed to it. Rather, the fluoride ion is the hazard. One may scoff, saying "Ha, drinking water has fluoride ion in it!" Right, but most water also has arsenic in it. In both cases, the issue is one of concentration. Fluoride ion makes for stronger bones and teeth because it reacts with calcium to form strong bonds; but it also binds to free calcium ions to form insoluble CaF2. In fact, too much fluoride in one's drinking water will make teeth and bones crumble.

The full story, according to Greenwood and Earnshaw's Chemistry of the Elements (Ref 3), is as follows p.946):

The highly corrosive nature of HF and aqueous hydrofluoric acid solutions have already been alluded to and great caution must be exercised in their handling. The salient feature of HF burns is the delayed onset of discomfit and the development of a characteristic white lesion that is excruciatingly painful. The progressive action of HF on skin is due to dehydration, low pH, and the specific toxic effect of high concentrations of fluoride ions: these remove Ca2+ from tissues as insoluble CaF2 and thereby delay healing; in addition the immobilization of Ca2+ results in a relative excess of K+ within the tissue, so that nerve stimulation ensues. Treatment of HF burns involves copious sluicing with water for at least 15 minutes followed by (a) immersion in (or application of wet packs of) cold MgSO4, or (b) subcutaneous injection of a 10% solution of calcium gluconate (which gives rapid relief from pain), or (c) surgical excision of the burn lesion. Medical attention is essential, even if the initial effects appear slight, because of the slow onset of the more serious medical symptoms.

Pragmatically, the single most important thing about working with HF is to take it very seriously and to not take a chance at not noting exposure. As mentioned above, HF is very slow-acting, and on contact, there is not an instant burning sensation, as with hydrochloric or sulfuric acids. By the time one feels a HF burn, it is far too late. Almost all superficial exposure to HF is minor if it is immediately noted and acted upon. If any amount of HF gets on the skin, rinse with copious amounts of water. This is usually enough if done immediately. An important exception: places that cannot be rinsed well. For example, the cuticles of the fingers and under the fingernails. Any exposure of such areas should get professional, and informed, medical attention. Do not assume most doctors know how to treat HF burns. Explain the medical effect of HF on tissues, have a copy of one of the references listed below, and bring them to the hospital or clinic. Two quotes from reference (1), below, illustrate the point:

Case 1: A drop of concentrated HF splashed on to the finger nail of a patient. The finger was insufficiently washed. The exposed point turned gradually to a white-yellowish colour, but no further visible changes were observed. Pain occurred 7 h later, which continued for about 30 hours. Examination of the tissue under the nail then showed that a pea-sized area had already been destroyed by necrosis requiring surgical treatment.

Case 2: HF had dropped on to the finger of a patient. Insufficient treatment resulted in the amputation of the exposed finger. Chemical analysis of this finger gave a fluoride content of as little as 0.16 mg.

Wearing gloves doesn't guarantee safety. Perspiration on the skin (induced by gloves) makes it difficult to know if a droplet of liquid on the skin is HF rather than sweat...and gloves do spring leaks eventually. Wash your hands thoroughly after removing the gloves, and change gloves at minimum every ten minutes when working with HF, rinsing the gloves before removing them and also washing and drying your hands thoroughly between gloves. Also change gloves every time you finish working with HF solutions, and are ready to move on to something else. If there is one thing worse than getting HF on yourself it is causing it to get on one of your co-workers, who will have no idea what is happening to them.

Robert Rossi

References: 1) "Symptons and treatment of HF injuries" D. Peters and R. Miethchen, J. of Fluorine Chemistry, Vol 79, pp. 161-165 (1996).
2) "Recommended medical treatment of HF acid exposure" Allied Signal HF Products Division. Morristown, NJ, 800-622-5002.
3) Chemistry of the Elements, N.N. Greenwood and A. Earnshaw, Oxford: Butterworth-Heinemann Ltd., (1984).