Scrubbing of chlorosilane 1

[SrChemE]

Hello,
We are in the process of designing a system to hold Hexyltrichlorosilane (HTCS), which is a flammable and combustible liquid. Vapors can form explosive mixture with air (it can liberate Hydrogen gases).
The "Safe Handling of Chlorosilanes" indicates that "operational vents from nitrogen blanketing systems should be directed to a vent recovery system or a scrubber".

Is anybody involved in a plant that handles HTCS or any chlorosilane? Our process will be installing a 24,000 gallon tank and we will need a scrubber.
If possible, I'd like information in designing this scrubber. I was unsuccessful in finding any material on line.

Thank you,
Sarah

 

[TiCl4]

Have you contacted the supplier of the HTCS? Vendors usually are quite willing to share material handling and storage design to customers, and they will have direct experience needed to give you specifics.

That said, I've worked with titanium tetrachloride, which also had small amounts of SiCl4. I've also separately worked with flammable gas scrubbing. I do not have any direct experience with HTCS, but I will give you my thoughts below.

Edit: Further reading makes me think you will not evolve hydrogen from HTCS. From GLOBAL SAFE HANDLING OF CHLOROSILANES (GSHCS), section 8.1: Hydrochlorosilanes, chlorosilanes containing at least one SiH bond, can generate hydrogen in addition to hydrogen chloride when exposed to moisture in the air.. HTCS has a Si that has 3 Cl bonds and 1 C bond. I do not see a Si-H bond.

If indeed hydrogen formation is not a by-product, the scrubbing system becomes simpler. I've edited out comments below regarding considerations around hydrogen due to this fact.

From what I've read briefly, HTCS behaves similarly to TiCl4 when in contact with water. Overall design considerations may include:

[ul]
[li]Low pressure N2 pad on the storage tanks to prevent moisture intrusion and keep O2 out.[/li]
[li]Venting during filling via pressure control valve. If HTCS forms scale or gel materials on contact with moisture, a pressure control valve is preferred over a spring-style vent because the PCV has the torque to power through any scale buildup. You can test for this in the lab by setting a sample out in contact with air (inside a hood) and see if any solids/gels form.[/li]
[li]There are many designs possible for the vent line, including placing the PCV at the inlet of the scrubber and maintaining the N2 pad from the tank all the way through the vent line. If gels/scale are an issue, the vent header should either be kept 100% dry or should have a water flush to carry formed gels through the vent. Avoid any low spots where liquid HTCS can accumulate. [/li]
[li]Any piping/vessel that is not controlled for moisture will likely need to be fiberglass or another material that is resistant to HCl. I'm not sure if fiberglass piping is resistant to HTCS, though.[/li]
[/ul]

Actually scrubbing the HTCS should be simple due to its high reactivity with water. If the vendor does not have a specific design, I suggest looking at acid gas scrubbers as a starting point for design - a small spray tower or packed bed should be sufficient for gas/liquid contact. We used a simple co and counter-current jet of water with caustic neutralization across an orifice to scrub TiCl4 vapors. Again, if solids/gels form, I would be wary of packing or other designs that are susceptible to plugging. The reactivity should be high enough that good scrubbing efficiency can likely be obtained at 7-8 pH. Sodium hydroxide is a somewhat cheap and easily handled neutralization agent for this. Other less expensive neutralization options, like slaked lime or soda ash, are a bit of a pain to handle, and probably not worth it for a small-scale scrubber.

Given HTCS is flammable, you can either keep the entire scrubber N2 purged (which may be expensive, depending on the N2 usage) or have a dilution air blower that keeps the HTCS content well below the LEL. If the latter is used, a downstream LEL sensor and explosion isolation valve or blowout panels are recommended.

You may want to consider putting a reflux condenser (preferably with a non-water based coolant) on the top of the storage tank to condense a majority of the HTCS before it leaves the tank. This would sharply decrease the required size of your downstream equipment and may be less expensive overall. If implemented, the condenser should be a double tubesheet design to prevent a tubesheet leak from entering the tank.

I hope that some of this may be helpful for initial design considerations. Whatever you end up doing, I strongly recommend a LOPA be conducted before you install this system.