Removal of Pd residues

[ndt228]

My problem is probably the opposite of most posts here. I do not want to recover Pd, but to get rid of it.
I work for a company that carries out chemical reactions in organic solvents, typically catalysed by Pd, as Pd(0) held in solution by a variety of complexing agents. Unfortunately over the course of a reaction some of the Pd metal comes out of solution and is deposited on the walls of the reactors - we know this as palladium black.
Our problem is that when moving to other chemistry in the reactor we need to clean off the deposit as (a) it could adversely affect the next reaction (b) it is a visual contaminant and not generally acceptable (c) it could leach off and become an unacceptable contaminant of the next product.
Current cleaning methods use nitric acid, or else the scouring action of granular salt suspended in an organic solvent like methanol. Obviously considerable care is needed to avoid any possible contact ! One issue is that our reactors are large (up to 1000 L) and thus to fill a reactor to ensure the inner surface is fully wetted uses a lot of material, which is both uneconomic & not very "green". The other (worse) problem is that these methods are not reliable and we get a lot of down-time for repeated cleaning attempts.
I read that Aqua Regia is more effective than plain nitric acid at dissolving Pd, but the classic 1:3 molar mix of conc Nitric to HCl seems even more hazardous than the (not really acceptable) nitric acid alone. I have also read of the use of a combination of thionyl chloride and dimethylformamide, but this carries other health risks due to the byproducts that form.
Does anyone have experience/suggestions that are relevant to this situation ?
Many Thanks
Nick

 

[ndt228]

To ask some specific questions:
1) Does anyone know if "weak" aqua regia (ie a small% of HCl in HNO3) might be any more effective than HNO3 alone ?
2) Might there be any other additive that makes (dilute) HNO3 more effective (we find that below 60% HNO3 it doesn't work well at all) ?
3) What minimum temperature would be required (general experience) with whatever reagent is used ?
4) Any other "magic" recipes ?

Thanks again
Nick

 

[JoeyJoystick]

Hi,

I can not help you with a chemical solution. However, worked with reactor vessels a lot in the past. From small ones of a couple of liters for laboratory use to large reactors in refineries. The cleaning of these vessels is at times problematic. And in some case this is where we were called in. So here is my 2 cents worth.

Considerations:
- Vessel 1.000l.
- Pd contaminations deposits inside the vessel.
- Green solution.
- Minimize waste.

Given this, I would not rinse the vessel. This would indeed require a large amount of liquid. I would modify the vessel to allow for the following. Allow for draining on the bottom. I am sure this is already possible, but if not, this would need to be looked into. And then I would make sure that I can insert a so called roto-head from the top. This is a tank cleaning nozzle that rotates when pressurised. This might require a modification to your set-up. However since it is a production set-up and you did mention reducing down time I think it is at least worth looking into.

Now these roto-heads are typically used for large set-ups and require huge pumps and very high pressure. But I do not think that this is required for you. It is a small vessel to start with. so the distance travelled by the yet is very short. For 1.000l vessel the longest distance travelled will probably be less than 1 meter. This helps to reduce the pressure significantly. The kind of fouling and previous cleaning attempts mentioned also indicates that a lower pressure may be sufficient to remove the Pd deposits.

So now you quickly rinse of the inside walls. You can than do a second run to neutralise the cleaning agent. And you than do a third run with water.

Roto-head.

This is a high pressure (not so high in this case...) cleaning nozzle which rotates when pressurised. All this while the yet from the nozzle will effectively reach all the inside of the reactor walls. Providing there are no obstacles inside the tank. We have used this in large crude tanks and sprayed into 0.5m thick sludge from 35m away. This would do miracles in terms of cleaning. For obvious reasons you can not use this in such a small vessel. But I think the concept may well work for you when scaled down.

Just out of curiosity though, what material is the vessel made of?

Joey

 

[snoman701]

Have you tried washing with formic acid first?

 

[Lou]

I would solve as follows:

Remove organic residues from tank.

Fill tank with 5 M HCl.

Introduce hydrogen peroxide as oxidant (1.5 mL 32 % wt H2O2 per liter of 5 M HCl).
Heat to 40-50 C for 1 h.
Cool to room temperature.
Add in 0.5 g sodium metabisulfite per liter solution. If you are able to reflux solution, the sulfite can be skipped.

Rinse tank until colorless.
Circulate Pd resin through IX resin for recovery.
Reuse acid.

 

[Geo]

Will chlorine gas dissolve the Pd powder without having to wet the entire surface? HCl and hypochlorite to generate chlorine gas should dissolve the solid metal. Chlorine gas is heavy and will fill the flask like a liquid as long as enough is generated. Then a water rinse to remove the solution. This may or may not work. It works on gold powder but I'm not sure it will work the same with Pd, but I suspect it would.

 

[Lou]

Sure,

You can reflux the HCl down the walls but you’ll have to keep adding in an oxidant to generate the chlorine.

1000 L reactor is 1 275 gallon IBC “tote”. It also makes the math easy. Every ppm in one of those is a gram of dissolved metal.

At 30 ppm it’s pretty visible as a light beer color. At 300 ppm, it’s getting dark and time to recover!

 

[ndt228]

Thanks for your collective inputs - all worthwhile food for thought.
To respond to some queries/pose further challenges:
a) It is an enamelled steel reactor (Pfaudler type tho perhaps another make), with an BRO valve
b) It has a white enamel lining which unfortunately makes any deposits very visible
c) I was already wanting to go over to a spray-ball approach. However this will require the recirculation, and whilst we are equipped to do this for our "normal" solutions in organic solvents, pumping HNO3 (etc) will be regarded as especially hazardous. Thus not done lightly.
d) Spray-ball material of construction ? The most common are Stainless Steel, but even though JNO3 is passivating under some conditions, once the chlorine is introduced, via HCl and/or peroxide, there's no way way we could use it. PTFE perhaps ?
e) I will ask around about formic acid - we already use it for other cleaning duties.

Thanks again
Nick

 

[snoman701]

Formic acid is just for reducing the PdO so it's easier to dissolve with HNO3 or HCl.

Lou's method is really going to be your most effective, and safest. Note that he was not using HNO3 at all, simply HCl with an oxidizer. Save the same tote of HCl and just use it for cleaning this process, then recover the Pd if and when it's worthwhile.

For intermittent use, you won't need something as inert as teflon for use with HCl, as polypropylene or cPVC is fine at a fraction of the cost.

If using concentrated nitric, teflon is going to be necessary if it's under pressure...but you can also use stainless, which is probably more suitable for your other processes. 304 is better than 316, as there is more chromium in 304.

One note, If you have stainless exhaust, I'm not sure how I'd feel about using HCl. I routinely use stainless equipment with HCl, but I also have the ability to quickly wipe it down with nitric immediately following use.

 

[JoeyJoystick]

c) I was already wanting to go over to a spray-ball approach. However this will require the recirculation, and whilst we are equipped to do this for our "normal" solutions in organic solvents, pumping HNO3 (etc) will be regarded as especially hazardous. Thus not done lightly.
d) Spray-ball material of construction ? The most common are Stainless Steel, but even though JNO3 is passivating under some conditions, once the chlorine is introduced, via HCl and/or peroxide, there's no way way we could use it. PTFE perhaps ?

Hi Nick,

If the reactor vessel has an enamel lining on the inside, you have a very rugged, chemical resistant vessel. The enamel doesn't like shock and may than chip away, but it has got very high chemical resistance but also a very high mechanical tolerance. You may want to consider jetting it off with high pressure water. You would still have to recirculate to remove the Pd but since this does not dissolve in water that should not be all that difficult. And you than do not have to work with chemicals for cleaning anymore. I would try cleaning at 100-300Bar. For the lower end of this you could do a trial run with a home high pressure cleaning unit such as Black & Decker. For the higher end of this range you could use a Karcher or something like that. These would still be kind of affordable considering it is a test and additionally, quite a few people/companies have them. If it is confirmed to work you can prepare for a proper set-up.

Do not underestimate the power of pressure. We once did a reactor modification in Singapore. Reactor diameter was about 2.5m Vessel was 25mm steel with a 50mm concrete lining on the inside. We attach the rail around the vessel and run the nozzle al the way around. Probably took some 20 hours. Surprise though. The jet had damaged the concrete on the opposite side of the vessel. This was due to the shear force of the High Pressure Water jet of course, but also because we cut the vessel under an angle to get the vessel baffled for easy and quick welding afterwards. OK, a different ball game, but all I am saying is that removing Pd from the inside of the vessel could be an option I think. Even when only using water.

If you confirm water as a cleaning agent works, then the final set-up will be made relatively straightforward since you do not have to consider chemical resistance. You can simply use a standard nozzle, seals etc.. Plus the Pd removal through circulation.

Due to the relatively high density of Pd you may even get away with separation though settlement if you have the space. as in; a large volume equals a low flow and allows the Pd to settle to the bottom. Alternatively, if you can confirm that the method works well for cleaning of the Pd from the inside of the reactor vessel, you could use 2 holding tanks. Pump clean water from one yank through the vessel to the other tank. This way you are sure you're not reintroducing any fowling back into the vessel. And the tanks are sized such that it will hold more than you would actually need for a cleaning run. And now you do not have circulation at all anymore and only introduce clean water into your system.

Joey

 

[Lou]

We clean our small Pfaudler in the fashion I outlined above. Occasionally we get a Pd coating in it if we are using reducing chemistry. Our glass is blue so less easy to see if clean.

The cleaning process could easily be controlled using UV-Vis. As coot saturation on a sample peaks no more Pd is dissolved.

 

[ndt228]

Once again,
Thanks for the further input.
I am happy to confirm that (in a controlled lab trial) adding 1 eq of oxidant (wrt amount of Pd charged) to 5M HCl makes for a quick and efficient dissolution. Further, I have been able to drop the HCl concentration significantly without loss of performance. I could readily prepare ca 5mg/ml Pd solutions in this way, which is well above anything I'd expect to be needed in real life.
Water jetting is something that has been done here in the past, and is still resorted to at times now. However, I have heard that on one occasion the recoil on the (Karcher) gun caused the lance tip to hit the vessel lid, and it chipped the enamel. Rule 1 - you just can't win !
I have found a variety of spray nozzles available in either PTFE or PVDF, and expect that we'll shortly be doing some trials..........once we have talked HSE through the risk/benefit of swapping from a large inventory of HNO3 to a small inventory of much more dilute chlorine.
Best regards
Nick

 

[Lou]

The chlorine can be removed with sulfite as mentioned above. You will probably inventory that material until you have enough to recover or send to a refinery.

FYI, I do accept those solutions at our refinery.

Best,