Wednesday, May 30, 2012

Process Wednesday: Carbon steel versus organic acids

From the pages of Chemical Engineering Magazine, an interesting story about an extraction unit (I think this is somehow related to a petroleum fractionating column?) from Mike Resetarits, the technical director of Fractionation Research, Inc.:
The natural state of the extraction solvent was clear and slightly yellow. With time, however, oxygen would creep into the unit, especially via the vacuum regenerator. Organic acids would form. The acids would eat any carbon steel that they encountered. Iron oxide and carbonaceous materials would accumulate in the solvent, which would turn from yellow to green to brown and then black. Eventually the solvent would look and feel like the three-year-old engine oil in my 1974 Ford Falcon.  
One troubleshooting visit was in response to a complaint about reduced recovery. The unit was shut down for the annual turnaround. My colleague, Reese, and I decided to enter and inspect the extractor at three manhole locations, starting with the top manhole. The top manhole was open. According to our drawings, the top tray was only about 3 ft below the top manhole. We inspected the top tray with our flashlights. It looked fine -- no fouling. Then Reese entered the column feet first, on his belly. He almost fell 20 feet.  
Unfortunately, Reese and I failed to expect the unexpected. The top tray was there -- but it wasn't. It was paper thin. It crumbled to dust instantly under his weight. Fortunately, he was wearing a harness with a lanyard. It turned out that the top ten trays were similarly thin. Replacement trays were provided to the plant on an emergency basis. 
One of the things that I'm encountering as a relatively new process chemist in a manufacturing environment is how much focus there is on the materials of construction of equipment. Most of the time in the lab, your compounds are only going to see one material: glass. But in the plant, glass is relatively rare, it seems, and concerns about the compatibility of whatever metal you might be using with your reaction solvent, reagents and byproducts becomes very important.


  1. Good job as usually, CJ. Keep posting about the less-than-obvious sides of chemistry. I hope more of us will read your writings.

    We like to view a lot of items in our lab lives as inert, stationary, almost eternal, while they are nothing like that. To name a few:

    - jobs
    - pretty things made of glass
    - shiny things made of metals
    - jobs
    - white slippery things made of Teflon
    - analytical results
    - jobs
    - disposable gloves
    - dust masks
    - did I include jobs?

    Destroy preconceptions daily - Taiichi Ohno

  2. In our place the development chemists had the job of checking the corrosiveness of reagents. It was part of the hand-over documents required when a process was moved into chemical production. In fact I think it was documented as a SOP.
    We could get these tests done in house, or latterly outsourced them as a package with the whole ecology testing.
    So if you are in doubt get a corrosion expert to examine your equipment or get him to test your reagents and reaction mixtures against steel, hasteloy etc.

  3. Glass is a great choice, but only for small-scale work. Making a 2000 ml 3-neck RBF is easy. Making a 2000 liter 3-neck RBF is impossible. The best you can hope for is a glass-lined tank, and those have other issues (differences in thermal expansivity for starters).

    The best example that I can think of with my polymeric mindset that shows the flexibility of glass at the bench is an all-glass extruder that I saw in an article once.

  4. 2000 liter is one big RBF. The largest RBF I worked with was a 200 L. Just like their 2 L cousins they like to float in baths, except a floating 200 L may push up against their glass overheads with 100+ kG of force. Also, heat transfer through the thick wall is pitiful.
    No, this wasn't my design.

    1. at this volume you would definitely need a jacketed reactor and heat transfer fluid circulator + efficient stirring. Thermal runaway can be pretty frightening even at a 2L scale...

  5. A friend I have worked for awhile in a hydrothermal chemical engineering lab. They studied chemical reactions in supercritical water (~400 C, hundreds of atmospheres). They seemed to use a lot of Inconel vessels or silicon-lined steel. One problem they worried about was how to tell if the reactor walls were catalyzing their reactions. Is that a concern for organic process chemistry too? If you can't use glass and you have to switch to steel, is it ever a worry that the metals will start catalyzing stuff?

    1. Typical concerns have been about incompatible reagents/solutions corroding the steel and putting non-ferrous metals into the solutions (e.g. chromium.)

      I haven't heard concerns about catalysis, but that doesn't meant that it can't happen.

  6. Certainly iron and other metals catalyze the decomposition of many compounds, one that comes to mine is an benzyl halide. You check this during the development by running the reagents, SMs and reaction mixtures in a DSC apparatus without the presence of metal then repeat using iron. Occasionally you will see a significant shift in the exotherm of decomposition, it can reach shifts of 80 - 100°C or more towards the low temperature side of things. (e.g. from 150°C to 70°C). This may take your reaction conditions into the danger zone where a runaway could occur if metal somehow got into your reaction. Don't forget that metal pipes and tubing are used in a plant for material transfer. I have used glass tubes, but it is a real pain in the ass as the glass may have to be specially fabricated.

  7. Glass has its own advantages, but also many disadvantages. The first plant I worked at had glass-lined reactors and all-glass overheads. Glass pipe is very fragile and the connection flanges have to be torqued to very precise (and low) torque. Even then, ageing gaskets kept us busy chasing leaks much of the time.

    Glass can corrode and spall and the damage is hard to notice. Recently there was a lot of work done on glass delamination in vials storing pharmaceuticals. More information about this can be found in under Alerts | Glass Delamination. There were several recalls of pharmaceuticals over this issue.

    Re: RBF vs. glass jacketed reactors - I actually don't particularly like jacketed glass reactors over 50 L (that 200 L RBF still gives me nightmares...). The heat transfer through the glass is so poor that heating/cooling rates in a 30 gal GL reactors can be 2x of those in a 100 L all-glass vessel.

    After taking into account the risk of accidental breaking the entire vessel with a dropped wrench or cracking the jacked when isolating the jacket without admitting an air pocket I would take a clumsily looking Pfaudler any time.
    If I need to have visual control beyond a manhole/flashlight I can use a PVM or a borescope.

  8. it must cost a lot of money to construct the jacketed reactor..

  9. Nice explanation.. For further more information We are engaged in bringing forth the clients an excellent assortment of Precision Machined Parts you can refer Precision Machined Parts


looks like Blogger doesn't work with anonymous comments from Chrome browsers at the moment - works in Microsoft Edge, or from Chrome with a Blogger account - sorry! CJ 3/21/20