A sensor-ble solution

A confession: I like cool stuff. I like products which solve problems. Being an analytical chemist, I particularly like products which solve problems in analytical chemistry. Not so long ago, I was shown a new development from Metrohm which I regard as just straight out, simply cool. It's an ion selective probe for calcium determinations, which uses replaceable, thick-film technology sensor tips.


 Now, as you will note from previous posts, my niche speciality is thermometric endpoint titrimetry (TET). In comparison with TET, I don't know all that much about measurement with potentiometric sensors outside of pH electrodes, and previous experiences with ISE's haven't been all that positive. However, this new approach by Metrohm seems to be a most pragmatic approach to the problem of limited service lives of polymer membrane ISE probes, and that is, simply replace just the sensing tip rather than the entire probe.

I guess that following through on this pragmatic approach, they've chosen a Ca-selective probe, as this could have considerable market appeal in the titrimetric analysis of water hardness. It will be interesting to see how the market accepts it, and if they follow this with tips which are selective for other ions. 

Goodbye to Gove (Part 1)

Late in 2013, Rio Tinto Alcan announced it would close down it's alumina refinery on the Gove peninsula of Australia's Northern Territory. I felt sad at this decision, although I understand the business decisions behind this. The reason for my sadness (as well as that of my wife) is that the years that we spent at Gove were some of the happiest of our lives.

Some background: alumina is the oxide of aluminium from which aluminium metal is smelted. The alumina is isolated as a pure compound from its chief ore, bauxite. Although alumina can be made from other aluminium sources such as clay, extraction from bauxite offers the easiest and most economical route. Bauxite can be thought of as what's left from a rock which contains aluminium after it has been subject to weathering over very long geological time periods under certain conditions of rainfall and temperature. The most economical bauxites to process tend to be found in tropical regions. Bauxites which contain the highest proportions of gibbsite, gamma-Al(OH)3 are the most highly prized, because they require less energy to process. The Gove peninsula is located only 11 degrees south of the equator, and the bauxite there has a high proportion of gibbsite.

Here's where Gove is located, relative to the rest of Australia:


It's a remote place. The original majority owner and technical manager of the mine and refinery Swiss Aluminium Ltd (or Alusuisse) made a documentary for their shareholders entitled "Keine Strasse Fuehrt Nach Gove" or "No Roads Lead to Gove", which was the truth. Everything that came into Gove came by sea via coastal freighter or barge or by air. The 4WD track out to Katherine was used by adventurous spirits travelling in convoy during the "dry" season (June to September).

For nearly six years, we called this place home, and we came to love it. My wife recalls that feeling of coming home as our plane bringing us back from leave flew over the bauxite mine, a hematite-red gash in the tropical woodlands.



As an amateur photographer, I loved the intensity of colour that the "wet" (monsoon) season brought, and I recall waiting for the bus to work in the mornings, gazing eastwards over the Arafura sea as the sun rose as a glowing orange ball, tingeing the gathering clouds which would later dump their cargo of rain in near-solid masses of water.

For me, there was fun to be had on two levels in Gove. There was the social aspect, where lots of mainly young people with young families came together. There were barbecues, dinner parties, and a lot of sport. In those early days at Gove, we made our own fun.

The second type of fun for me were the technical challenges. Although by profession I am a chemist, for the duration of the start-up of the plant I worked as a process engineer, responsible for starting up one of the unit processes of the plant. After a couple of years, I was asked to take over management of the process control laboratory. 

So, how can you have fun in a place like this?

The first real challenge had to do with analytical chemistry, more specifically the analysis of the process liquor circulating in the plant. Let me explain the significance. The process used in alumina refineries around the world is called the Bayer Process, name after the Austrian chemist Karl Josef Bayer, who invented the alkaline route to the production of alumina in 1888. It's really a very simple process in principle. Bauxite is an ore containing mainly aluminium hydroxide and oxiyhydroxides, iron and titanium oxides, and silica in the form of kaolinite and quartz (fellow ex- and present alumina workers, please forgive this and following over-simplifications). The alumina values are separated from the other minerals by dissolving them with caustic alkali to form a supersaturated aluminate solution. After separation from the "red mud" gangue, the clear supersaturated aluminate solution is cooled and seeded with Al(OH)3. This causes some of the aluminate content of the liquor to decompose to Al(OH)3, which is then filtered, washed, and calcined at high temperature to alumina, Al2O3. The aluminate-depleted liquor is then recycled to dissolve more aluminium values from incoming bauxite. I'll talk about this in more detail in a later post. Essentially, the Bayer Process can be described in the reversible reaction:

Al(OH)3 + OH(-) <> Al(OH)4(-)

Our problem was in the manual, indicator-based titration method gifted us by the technology provider of the plant. Every operator-analyst saw the endpoint differently, and to add to the misery, the reaction behind the titration wasn't stoichiometric. The net effect that the production department didn't trust the results the lab. was issuing, and wouldn't run the risk of driving the process to the point where it was most efficient and where the profits are made.

A predecessor of mine had purchased, but never put the effort into implementing an early type of thermometric titrator. I saw it as our last best hope of solving the problem. Given that all my lab. supervisors were flat out with their daily tasks, it fell to me to do the development work necessary. Being a stubborn fellow, I kept at it until I felt I had a viable method. Then came the really difficult bit. We had to prove that it would actually control the process. We did lab. experiments on mass balances, and then tracked the process in parallel with the existing liquor analysis procedure. We proved we could predict the bauxite charge to within 1 tonne in 150, while the existing procedure couldn't close mass balances, and couldn't predict anything worthwhile. Still, we had to overcome opposition and even some hostility from the production and process engineers. I was later reminded of something the great Italian management advisor Niccolo Machiavelli wrote in his work "The Prince": "It should be borne in mind that there is nothing more difficult to handle, more doubtful of success and more dangerous to carry out than initiating changes (in a state's constitution). The innovator makes enemies of all those who prospered under the old order, and only lukewarm support is forthcoming from those who would prosper under the new". Substitute "analytical method" for "state's constitution" and you get my drift.

We persevered, and unwittingly at the time, provided the analytical tool that would permit the successful implementation of a radical change in the process necessitated by the world demand for a change in the type of alumina required by modern smelters. I'll talk about the laboratory's role in this new process in a later blog. 

There are a couple of footnotes. Firstly, I was told some years later that the change in the analysis procedure had been responsible for the production of an extra 50,000 tonnes of alumina per annum with no additional process costs. At a then price per tonne of approximately US$ 200, I'll let you do the math on the positive benefit to the bottom line of the company. Secondly, this positive experience with thermometric titrimetry led me years later to co-develop a modern, computer-driven automated thermometric titration system, whose enabling technology is now incorporated in the Metrohm 859 Titrotherm.

Where have all the chemists gone?

A long while ago, the folk singer Pete Seeger wrote a song "Where have all the flowers gone?" It was a favourite of the pretty, soulful girls with ironed-straight long hair who sang in the coffee houses of my youth.  In my work with industry over the past years, I could almost sing to the same tune "where have all the chemists gone?". Where indeed, for they seem to be few and far between in many quality control laboratories these days.

For those of us at the pointy end of customer contact in the analytical instrumentation game, it's becoming increasingly difficult to find people who have the technical knowledge to understand and successfully operate relatively simple instruments. I have personal experience of a large food company with a factory in my home city where there is no analytical chemist on the staff. I was involved in the installation of a titration system and training when it was sold two years ago. Our point of contact is the "technician" who is supposed to provide in-house support to the factory workers who do the actual analyses. The "technician" is actually a terribly nice bloke, but with absolutely no background in chemistry whatsoever. Over the past two years, he has been in frequent contact with the sales staff, expecting them to sort out simple problems which are principally of his own making. In another case, I did the installation of another system in a chemical packaging company where there are no chemists on the staff. They thought that it was perfectly reasonable to expect that included in the purchasing price was an ongoing commitment to sort out their problems as they occur. I stress that overwhelmingly, these problems stem not from equipment or component failure, but purely from a lack of understanding as to what they are doing. I feel a suitable analogy would be walking into a car dealership and expecting the salesperson to teach you how to drive as part of the purchase price of the car. A colleague recounted a similar experience of his, where the previous job of the laboratory supervisor of a factory belonging to his customer (a large chemical concern) was  as a pastry cook. His unqualified laboratory staff have reportedly appallingly bad analytical habits, and resolutely refused to listen to any advice as to how do things properly. It is a fact that unqualified persons can be trained to push the buttons on modern automated analytical instrumentation, but there has to be someone somewhere in the company with the requisite qualifications, experience and knowledge to fix the problems when and if they occur.

Is it reasonable to expect that companies outsource their technical expertise to instrument companies who may have little knowledge of the processes or products? There was a time when development chemists were common in analytical laboratories. There was a realization that improvements in analytical methods lead to better process control, better raw material utilization, and less wastage, all activities which drive profits to the bottom line of a company. So what has happened to bring us to this point? Your comments, please.

Here I am again....

It has been a long while between posts. More than six years, actually. I started this blog with the intention of promoting the little-known technique of thermometric titrimetry. However, instead of just talking about it, I have been involved in getting it installed in process and quality control labs around the world. After my company sold our technology to Metrohm AG, I have been supporting their marketing and sales efforts to make it a part of that company's palette of titration techniques. It's been a struggle, but I think that Metrohm is eventually getting there. In itself, that's quite an achievement. A number of companies have tried, but until now, none have succeeded in making the technique a commercial reality. 

Along the way, I've developed a range of applications for thermometric titrimetry across many industries, and even some which are novel to the industry. I've written nearly 130 application notes and countless customer reports; as well as replying to countless emails from customers, applications chemists and salespeople. I've installed instruments and trained customers and Metrohm applications and sales staff all around the world. Sometimes it's been frustrating, and often there's been a lot of hard work, but when a method comes together and you've solved a customer's problem, there's a lot of satisfaction. In most cases, it's also been fun.

Now it's time to kick back a little, and take it a little easier. I'm still working on new applications, but at the moment confining myself to helping local colleagues. I've also got a little time to think about other things, and perhaps post some of these ramblings here. Let's see what develops.

So how does this thermometric titration stuff work?

So how does this thermometric titration thingy work? Well, labrat old son, I'm glad you asked that question. You can see how desparate things get around here when I have to answer my own questions
The answer is: exactly like any other titration, except that the heat of the reaction between the analyte and the titrant is used to detect the endpoint. Let's take an example: you want to determine the sulfate content of a particular sample. Your titrant will be 1 mol/L barium chloride. The reaction of barium ions with sulfate ions to form insoluble barium sulfate is exothermic (heat is given out by the reaction). In a thermometric titration, the burette delivers titrant at a constant dose rate, so the reaction between the barium and sulfate ions is occurring at a constant rate. In an exothermic reaction, you'd expect that the temperature would increase at a constant rate too. There are a few other things around like heats of mixing and stuff, but by and large, we see an increase in temperature. Check the titration plot from a Metrohm 859 Titrotherm instrument below. The red line is the temperature trace of the reaction. The volume of titrant added is plotted on the "x"-axis, the solution temperature is plotted on the "y"-axis. When all the sulfate has reacted, there is nothing to increase the solution temperature at the same rate, and we see a break in the curve. That's the "endpoint" or "breakpoint" of the titration. We measure the amount of titrant added to this point, and calculate the equivalent amount of sulfate that this represents.

Finding the exact endpoint used to be a tedious business until cheap PC's came along to run the powerful algorithms needed to get accurate results. In thermometric titration, the second derivative of the temperature plot is used to locate the endpoint, as you can see in the next titration plot.The yellow curve represents the second derivative. The endpoint is at the peak shown on this curve. Of course, there's some pretty flash maths going on the background, but you and I don't have to worry about that. The results can be sent automatically to a dedicated spreadsheet for calulation, which all makes it pretty easy.

A Tale of Old El Paso (Lite)

There is a mineral called "elpasolite", because it was first identified in a deposit near El Paso, Texas. Sorry about the pun in the title, BTW. It's certainly nothing remarkable to look at. It's a sort of dull grey, not much lustre, and quite soft. You're not about to rush into your local jeweller and ask him to mount it into a ring for your significant other. So why is it my favourite mineral at this time? Well, it has the molecular formula NaK2AlF6. OK, that's confirmed your suspicion that I may be certifiably strange. But wait, there's more.

Elpasolite is the key to a neat series of thermometric titrations where you can determine sodium, potassium, aluminium and fluoride. The reaction is simply:

The reaction is exothermic, and proceeds at room temperature. This makes it a good candidate for thermometric titrimetry. By suitable arrangement of the experimental conditions, you can employ the same chemistry to analyze for Na, K, Al and F.

Let's take aluminium as our first example. Aluminium-based chemicals have important uses in water treatment and as anti-perspirants, among other applications. Aluminium in such materials can be tedious to analyze by titration. The traditional method has been to add an excess of standard EDTA, boil to ensure full complexation with the EDTA, then titrate the excess EDTA with Zn solution to a visually-indicated endpoint. It's not an easy endpoint to pick.

By contrast, the elpasolite aluminium method is easy. Just measure the Al containing sample directly into the titration vessel, add a buffer, and titrate with standard sodium fluoride. The buffer comprises sodium and potassium acetate and acetic acid, and brings the solution to ~pH4.5. The buffer is designed to supply the excess Na and K ions required to drive the reaction strongly to the right. The only proviso is that the Al has to be present as Al+++ . This means that for aluminium chlorohydrate, this material needs to be hydrolyzed with some HCl prior to the buffering step. Similarly, sodium aluminate needs to be strongly acidified.

The elpasolite aluminium titration is very precise, and analytical precisions <0.1%.

I first came across it in one of the texts that I have on thermometric titration, G.A. Vaughan, Thermometric and Enthalpimetric Titrimetry, Van Nostrand Reinhold Company (1973) The original method was for the determination of sodium by direct-injection enthalpimetry(!) and the reference given is: Sajo, I., Magy. Kem. Folyoirat 75 1-3 (1969). Since I don't know Hungarian, and since I don't have the gear for that type of determination, I jiggered around until I got it to work as a thermometric titration. It works a treat, too. Sharp endpoint, great precision, and fast as a whippet. Generally, you're looking at under 2 minutes for a titration.

Interesting, you say, but why would you use it when there's AAS and ICP around. Well....

1. The method is intended for Na in the g/L or % region, so with AAS and ICP you have to dilute and dilute and dilute and dilute and dilute until you get it in range with the instrument. With the elpasolite method, you just weigh your sample directly into the titration vessel.

2. You don't have to filter your sample, either, because with thermometric titration, you can titrate with solids in the sample solution.

3. In common with all titration techniques, it's a linear method. By that I mean that instrument response (measured in mL of titrant) is directly proportional to the amount of analyte present, whereas with spectroscopic procedures such as AAS and ICP, the signal is proportional to the logarithm of the analyte concentration, which is not so great when you're going after maximum precision.

Introducing me.

This is me. The labrat in the white coat. I've been a lab rat for a long time. Ever since I was small, I wanted to wear a white coat, and stuff the top pocket with spatulas, indicator paper, and things which you use to prod and poke. I can't seem to get out of the lab. Every time I try to get out, it keeps dragging me back (where have I heard that before?). Actually, I don't want to get out of the lab. It's where I belong.

So about this blog. This is going to be about what I like to do, and what I do for a living, which is pretty much the same thing. You can shout "get a life!", but all I can say is that I have a life, and this is it. I'm a professional analytical chemist specializing in the development of analytical methods for process and quality control. My company consults to the global market leader in titration technology, Metrohm. We are assisting them introduce the technique of thermometric titrimetry to the marketplace in the shape of their new thermometric titration system Titrotherm. So yes, this blog does have a commercial aspect to it, but I am a professional chemist and I do take the responsibilities that go with that seriously. That means that I will only make claims which can be backed up by published literature; either in the shape of application notes published by Metrohm or published elsewhere in scientific literature.