Dear all,

who has experience with amperometric detection of carbohydrates? We have a Dionex detector and we canīt find suitable conditions for the detection of traces of cyclodextrins.

Thank you very much in advance!

I am quite familiar with the technique; just taught a class on the subject last week. I haven't tried cyclodextrins, but I think I recall lit. references on the subject. What are your conditions, mobile pahse, post columns addition, pulse sequence. Detection of oligosaccharides is routine, so I do not see any reason that cyclodextrins would not give a signal. Are you sure you are eluting them? If you aren't sure, take out the column and do a flow injection of a modest concentration and see if you get a signal.

Dear Bill,

we are sure that our cyclodextrin elutes from the column. We use a Dionex AS11-HC and NaOH, c = 25 mM, as mobile phase. We do not use a make-up (i.e. post column addition) solution. The amperometric conditions we found after some trials are 0.1 V (we integrate the signal at this voltage for 200 ms).

We do not know how to optimize the amperometric conditions without measuring a hydrodynamic voltamogramm. Can you give us some advice?

Many thanks in advance!

sorry for the delay as I have been away...fixing a daughters roof.

The Dionex technical note 21 available from their web site provides pulse sequences for carbohydrate analysis. The newest sequence that uses reductive cleaning seems to be quite successful and I would recommend it for most work. The advantages of various pulse sequences are detailed in the technical note.

It is not easy to optimize a pulse sequence. Witness the fact that after 20 years Johnson, its inventor, came up with a greatly imporved one. Early on people optimized signal strength and experiments at hydrodynamic electrodes provided guidance. Later, it was discovered it was better to optimize signal to noise. As far as i can tell this is an empirical adventure. Later still, people found that long term drift of response factor needed to be considered in the optimization. Optimization involves 4 times and 3 voltages, at least, with three responses-signal, noise and long term stability. It is a boggleing amount of work to do an optimization and I have been satisfied to let Dionex do it. If you don't get an acceptable signal using one of the three recommended sequences something is amiss besides the pulse sequence program.

Detection involves some very special sites at the gold surface. The pulse sequences in Technical Note 21 establish a reproducible concentration of these sites. You didn't tell me what the rest of your program was. If it isn't right, you won't have the catalytic sites to do the oxidation. One can't deviate much from recommended voltages.

Have you tried a simple sugar? If it doesn't give a proper response something in the detector or pulse sequesne is wrong. A bad reference electrode can result in poor response. A crapped up counter electrode also can.

I never asked, but what concentration/amount are you trying to detect? Cyclodextrins are big molecules with pooky diffusion coefficients, so don't expect the response you might get for glucose.

Another point worth mentioning is that the AS11 column is not a good choice for the separation of carbohydrates. Columns designed and developed for the analysis of inorganic anions and organic acids with hydroxide eluent are generally not useful for retention of carbohydrates except in the special case when the carbohydrates have a high net charge such as polyuronic acids. Because retention of carbohydrates requires a high stationary phase pH, this generally necessitates the use of columns which are not hydroxide selective (and thus are not generally useful for inorganic anions and organic acids). A better choice would be the CarboPac PA1 column.