all figures
COPYRIGHT
Inst.f.Chromatography

Trace analysis by Micro-Planar-Chromatography

This chapter needs so much additional experimental and error controlling work, that it can be written only in parts over the next months. Therefore by now here

PART ONE:
Water Trace Analysis, checking drinking water quality.

This most important foodstuff comes into increasing danger globally. It is not so much polluted by materials coming from agriculture overactivity but by more and more medical substances and their metabolics through the waste water channels into river and ground water resources.
It now looks that the standard processes of water quality control by HPLC-MS and similar concepts concentrating on a few hundred known key impurities in groundwater, river water and other sources for drinking water is no longer acceptable. Recycled medicine impurities in water cover a spectrum of several ten thousand substances steadily growing in their individual concentration. Therefore positive single substance identification after complete separation is no longer possible. It is well known, that isomers may show very differing levels of toxicity up to cancerogenic effects so that critical toxicity analyses would need total separations down to the structure isomer. We should not forget: although never all isomers coexist in nature, just only the C32H66  hydrocarbon can have 27.711.253.769 isomers. (Andersson, J., at
Th.Welsch-Good-Bye-Symposium, University Ulm, Oct. 9, 2009).
See also www.mathe2.uni-bayreuth.de/sascha/oeis/alkane.html . Some chromatographers believe in total separation - a systematic error in thinking. More than seven lines reporting about substance and concentration data are the upper level for human decision making. This is the reason why the author proposes a completely different quality check concept. Unfortunately it only ends up with the ”NO - do not drink - continue clean up” warning. But this important alarm result is available quick, with a minimum of instrumentation effort, everywhere available, most economically and with a high level of analytical security.

Concept and technical details:

* We use the µ-PLC technique published in this Open Access book. We add a circular sample bottle holder and a strong cold air blowing hair dryer. We need an especially permeable wick in the 5 mm o.d. PTFE tube. The tube is together with a PTFE sealing ring a compact part for top hardware purity conditions, see figure 1. Thus the sample is not in contact with the screw top. A one ml water sample is transferred by a clean pipette into a clean 1.5 ml micro bottle. The bottle is closed with the above mentioned compact PTFE sealing disk compactly connected with the 5 mm o.d. sample - and mobile phase transfer tube plus wick. A figure of the PTFE sealing ring will be shown at the end of this part 1 soon.

* The water sample flows from a 1.5 ml micro flask through the wick accurately into the center of a top pre cleaned HPTLC plate of 100 x 100 mm size at room temperature.

* The PTFE tube does not directly touch the layer. A  few hundred µm thick wick part looking out of the tube end assures the sample flow but avoids any mechanical plate layer damage. NOTE: the touch pressure is given only by the weight of the sample micro bottle - about 4.5 grams. The wick itself has only 33 mg weight corresponding to a 20 times 11 mm piece of viscose felt fabric area.

* The wick permeability is adjusted to a water sample transfer flow of about 50 µl / min.  Water samples are so viscous that well permeable wicks are necessary. The permeability is measured by the back pressure of a 2 liter per min air flow from the µ-PLC gas pump. How this is done and for quantitative data see under “Making a µ-PLC instrument”.

* Air from a clean air room is blown by a strong hair dryer with removed heating wires at full speed onto the open plate surface with the blower tube symmetrically adjusted above the sample bottle in about 200 mm distance from top - see photo at the end of part 1. The heat generated by the blower motor warms up the blowing air from room temperature by about 2 to 3  degree centigrade.

* After about 20 minutes we see the one ml water sample now only as a wet circular area of about 20 mm diameter - this part of the sample water is much colder than the blowing air because of the evaporation energy taken. Organic water impurities remain sharp at the outer border of the tube holding ring. When the sample bottle is empty  it is taken out of the tube holder ring and the wet sample area is dried to completeness by the still blowing dryer.

* Now the sample bottle is filled with one ml freshly distilled cleanest methanol. Under still flowing dryer air the bottle is reinserted into the center PTFE holder ring. Because of the lower viscosity the methanol flow is now larger than 50 µl/min, thus the residual sample parts and all remaining traces in the transfer wick are quicker transferred  into the circular area of the adsorbed water impurity substances. The transmitting methanol dries off sharply outside the PTFE bottle holder ring, focussing possible substances from the water sample.

* Depending on the sample impurity level we can transfer instead of a 100% water sample mixtures of water with volatile solvents like methanol, ethanol, butanol, acetone, diethyl ether. This lowers of course the best possible detection / quantitation level but may avoid chemisorption of impurities on glass and/or the wick material.

* The next steps are standard µ-PLC procedures checking
- the sharpness and position of the focussed sample start circle,
- the detectability conditions,
- the separation procedure for very first raw group separations -
         like highly non polar - highly polar,
- photo documentation,
- add-on of known quantities of external standards, locally brush sampled inside
         selected positions of the first focussed sample circle and refocussed onto the final
         water sample circle prior separation steps,
- final photo transfer to the multi-integration procedure

                       BUT if necessary or advisable only.

It is NOT advisable under the following conditions:

When checking the water quality as DRINKING water we may find on silica gel chemisorbed materials. They fluoresc and may not be movable by any solvent out of the about 5 mm large wick-to-layer touch area in the plate center. We may find under UV and under FLUORESCENCE conditions detectable impurities as a strong more or less thick substance circle at the former outer border of the bottle holder ring.
In this case further analytical effort is useless.
Because: if we see ANYTHING on the chromatogram area AND got before and after the water analysis the necessary critical BLIND run with zero signal results, the tested water is not of drinking quality.

We do NOT need any chemical identification, no MS/MS effort or what so ever.
This “No need for further effort” analysis is fast, economical and safe. It is based on the main condition of drinking water: it MUST be free from anything above the detectability limit  given by µ-PLC with UV-signal control. This is the ppb/ppt-level for organic substances.

However we may use such “No Drinking Water” chromatogram data to support clean-up-techniques and installations. In this case it may be necessary to identify parts of the signals. They may correlate with repairable problems in the clean-up process and thus help to identify error sources like not yet correctly cleaned pumps or other hardware parts.

If the drinking water data found by the water-µ-PLC concept given here  show ZERO substance signals we however need further effort, because we cannot state, the water is OK.
Because now we may be confronted with possible problems or systematic analytical errors:

> The water is possibly not of drinking quality because of bio impurities
    (bacteria and much more). This type of impurities has to be checked always in addition
    after high tech filtration and micro UV light treatment.
> Unknown substances may chemisorb on glass when transferred into the micro bottle or
    later on the transfer wick material when flowing into the layer.
    As the sample bottle and the wick however are flushed by volatile solvents after the
    water transfer into the layer and the flushed materials are focussed inside the first
    sampling area this error might not be of large impact. But who knows.
> The impurities cannot be detected - by NON of the up to hundred HPTLC detection
    reactions including gas reactions after the final focussing step.
    But still critical substances may pollute the drinking water.
> The detection limits are much larger for up to now not detected impurities than the
    ppb/ppt values basically possible when one gram sample concentrates in a narrow
    sharp layer area. Therefore other analytical techniques are still necessary.

Modern analytical development in the water analysis area is more and more directed to  micro-enrichment techniques (as an example: W. DUENGES: 20 ml water sample based micro enrichment procedures) prior HPLC separation online MS and other detection systems.

Quantitation limits:

If water has by chemical structure known detectable impurities - special organic traces - visible under UV by absorption or by fluorescence emission - and if the material amount of those traces is in the range of a few nano gram concentration per one gram of water sample, the quantition limit of these substances reaches 10 to the power of -7 weight-percent, which represents the ppb-level. However further effort depends on the special job to be done. If it is only the drinking water test, we do not need quantitation as discussed above. If the µ-PLC is taken to develop or improve clean up processes, quantitation is needed and easily possible by multi integration of the digital photo data. Quantitation  problems however remain in case the impurities are multi substance mixtures of unknown structure like original medical residuals and metabolic substances.
See here .

Detectability limits:

For the simple drinking water test the use of an UV-lamp and light filtered photography is completely sufficient. If we see strong signals, the “Do not drink it” decision is mandatory. For special jobs like mentioned above we may need other detection modes easily applicable to the 100 x 100 mm plate with central circular signals. This includes MS. But as gas reactions may also lower the detectability limits it is a simple task with the µ-PLC concept. The plate - aluminium foil based - can be heated up on a hot area, is covered with the standard glass plate with central gas inlet open to all sides and the gas stream flowing from the aquarium pump can transfer the reaction chemicals. This is easily possible under inert gases due to the flatness of the “instrument”.
 

At the end of this “part 1” - µ-PLC trace analysis :

it may look like the concept is analytically quite weak. One however must realize, that the state of the art of water analysis by regulated standard techniques is critical with respect to the 100 % rate of knowledge: All enrichment techniques used up to now fail with respect to completeness besides low levels of yield. Highly polar traces with chemisorption character we have seen already at the very first experiments are not known by now in standard techniques. They are not extractable by solvents but only when using solid adsorption materials. And how to detect them when fixed on solids ? The standard water analyst may now know what to do. Taking the water sample as such for enrichment directly into the analytical system is promising but the first results were disappointing with respect to the drinking water quality. Especially shocking were some results found with water samples from bathes in connection with medical care institutions.

Sampling of a mix of 0.5 ml clean methanol and 0,5 ml water sample needs about 29 minutes time under constant hair dryer action. When all sample entered the plate layer the bottle and its holder are removed. The still wet central part of the sample area is dried completely with the air blower still on. It blows all the time a strong enough air stream at 20 degree centigrade onto the plate.  When dry the plate is covered with the standard cover glass plate - see figure 4. The air blower is switched off now. The sample area is focussed with - normally - freshly distilled methanol as mentioned above and again dried with the cover glass plate taken off. Figure 5 shows a very first result for a (quite) clean water sample under fluorescence.

Analytical details to rain-, snow-, tap-, fresh and very old bottle-water, ground-, surface-, river-water will follow. There have been found new possibilities to change the selectivity.
(State of Sept. 2012)

PART TWO:

Specific Trace Analyses. 

First experiments with fruit liquids, as an example wine, look promising. We learned, that drying steps in between are necessarily to be done until total uniformity of the whole plate. The two figures below show drying errors resulting in separation selectivity effects - see the circle deformation. Sample amount into the plate center was 0.50 ml, drying time to get rid of the more than 85 % water in the wine was only 30 minutes. The enriched wine constituents were concentrated on a sharp circle as similar to figure 5 above.
The first separation run was made with methylene chlorid 97 / methanol 3 and a second one with methylene chlorid 90 / methanol 10  v/v.
Only one wine sample was taken which would not allow for critical sample comparison. Thus the enriching sampling must allow for 2 to 6 samples simultaneously for which we found a technical solution.
Follow up separation must allow for specific identifications and for calibration, which also was found to be a complicated task.
But note: the following two figures show only the same chromatogram under UV 254 and FLU 363 after only two conesecutive separation runs. What we want to demonstrate here is the sharpness of signal circles although we sampled 500 micro liter, a factor 100 above sample volumes used up to now in classical PLC analyses.

Lots of detailed studies however are necessarily to be done.

scroll bar -->