Soil Solution Sampling and Water Sampling Filtration

September 24, 2014

I wanted to share with you a conversation that took place a few years ago but is still as relevant today as it was then on the relative merits of analyses on soil solutions collected by ceramic or polysulfon plastics (Rhizon) which have a 0.1 micron pore size.

Take a look and please feel free to comment …..

There is a requirement to filter samples on site for all metals analysis. This is done with a .45 micron filter. The filtration is a requirement for METALS only and other analyses must be made WITHOUT filtration as some compounds may form colloids which are too large to pass through this pore size.

Dissolved metals are present as ions (i.e Zn2+), dissotiation of the salts. How much dissolves, depends on the solability product of the salt and the pH (also on pressure and temperature but these are pretty much constant in soils).  

Sampling with Rhizons is not a problem at all. The problem is sampling with porous material on which the metal ions can adsorb (metal lattice, or ordinary ceramics).

In relation specifically to the Rhizon samplers:

Leaving colloids aside, when metals dissolve in water are they present as atoms or as molecules or both. At atomic level anything will pass through 0.1 micron holes but if the metal is also present as part of a molecule; are most molecules small enough to pass through a 0.1 micron slot?

First TOC and pesticides are large non-polar molecules. They hate water and love to stick together as colloids, and won’t pass micro-pores as described. Therefore sampling TOC with soil moisture samplers is not a good idea, of course 1 or 2 molecules always will pass but reprecentabillity will be poor and volatiles are lost forever. You need a monitoring well to sample these compounds with alternative discussions about micro-purging turbidity also.

All the radioactive elements – minus tritium – are metals. These should be sampled as metals, meaning a 45 micron filtration. A bad or no filtration at all – will over-estimate the real concentration. Why is this interesting, because at present there is a huge discussion about measuring metals in soil? Some scientist will argue that for example mercury-sulphate is so stable it won’t dissolve at all and stick to the soil matrix for the next billion years. However during analyses all the mercury is dissolved and there is no distinction between the oxidation state of the element.

My problem is tritium, this is a waste product for nuclear plants, it is a hydrogen atom with 1 proton and 2 neutrons. Yes it can be found in hydrogen gas (H2), but why not in water H2O, with the heavy hydrogen atom or any other non-organic compound which contains hydrogen? Next question is then what do you analyse in the lab??

That is something we’re trying to found out but again maybe we’re completely wrong and tritium is only found in hydrogen gas (H2) and so instable it directly falls back into a more stable isotope and emitting radioactive radiation.

A straw punctured with many, many holes of 0,1 (or a bit more in) microns would work like a sampler with which you could take water from the unsaturated zone of the soil (Rhizons and fine porous ceramics work like that). If the holes are above 1 micron (by heart) applying a vacuum will cause air to pass instead of retaining it and only letting pass water. In other words; a straw punctured with “large” holes will operate like a piezometer; it will collect water only from the saturated zone (and a vacuum does not help). Sorry for the scientist.

Then the TOC and elements question: After establishing an equilibrium, the mentioned elements will pass. Dissolved tritium will pass also, except if the vacuum will de-gas it from the sample (with limited under-pressure the recovery on tritium would be higher). Other elements, even the large uranium molecules, are far smaller than the 0,1 micron of the polysulfon plastic. However when adhered to colloids (like tannin which is present in peat soils) they may not pass as the colloids surpass the diameter of the pores. Same counts for pesticides and similar hydrophobic organics like PAH. Their state of emulsion, rather than completely dissolved state, will prevent them from passing through the Rhizon membranes. This in fact is, I believe, much underestimated.

Following a site visit with some researchers from the University of Manchester. We were discussing the use of Rhizon samplers for their experiments and they told me they had experienced a lot of problems collecting dissolved organic carbon (DOC) after filtration through a .45 micron filter and the Rhizons, from literature have .15 micron (actually this is from the Eijkelkamp catalogue but their website says .1 micron as do the instructions for the macro rhizons?) pore size.

As far as we understand it DOC is defined as any carbon molecules which will pass through a 45 micron filter. The Rhizon samplers filter to 10/15 micron, and therefore a proportion of the carbon molecules would not be sampled, what proportion of the DOC this represented and whether this was a significant proportion? Again, possibly specifically for peat lands, organic matter will humify producing large long chain carbon molecules such as humic acid, tannins etc. therefore what fraction of DOC is made up of these large chain carbon molecules, but would suspect that there is likely to be a greater proportion of longer chain molecules comprising DOC than smaller carbon molecules. Additionally, on these samples people look at E4/E6 ratios which is dependent upon an absorption of specific wavelengths (465 and 665nm), would this bias the samples by reducing the heavier fraction of the carbon molecules, making the samples appear more or less humified, from which people infer the age of the sample water?

About the metals question: if the metal is in solution there would be no problem using 0.1 micron filter. This would also probably be true for nutrients such as Nitrate etc. However, peat lands may be more complicated than this because of the low pH values and the large amount of organic carbon molecules which can be extremely large in size, and would tend to provide adsorption sites for metals. If the sample is filtered to 10 micron then, like the DOC, a proportion of the large chain carbon molecules would be lost along with any adsorbed metals.

For peat land soils, these filters (10 and 45 micron filter) are more prone to clogging because of the amount of suspended material and may actually be sampling an even lighter fraction of the DOC?

It may be interesting to know at this point and time that I have studied the work (to some extend) of Mr. John McCarthy. I guess Mr. Baines will know his name as John McCarthy did a lot of study on the migration of colloids and adhered material, primarily in the unsaturated zone in the nineties.

It is his (and others) knowledge that led us (with me as strongest pusher/puller in this project) to a strongly revised Dutch standard on groundwater sampling (via monitoring wells).

We found out that turbid groundwater samples could exaggerate PAH, pesticides (etc!) content with a factor 1000 and more, compared to a clear (but not filtered) sample. When filtered the concentration of these low soluble components was reduced far below the really mobile fraction. By applying a sampling technique that reduces forces on soil particles in the formation around the screened part of the well this problem can be tackled.

This all led to a new groundwater sampling standard incorporating the obligation to reduce the force on soil particles around the screened section of a well by reducing the drawdown of the water level in the well (to max 50 cm). At the same time the resulting turbidity (which becomes in fact much lower than when e.g. a bailer is used for sampling) must be measured and noted. This gives the manager a tool to estimate the quality of the resulting analysis results. Natural turbidity is max 10 NTU under natural (gravity) forces.

The Americans apply the low flow sampling method already since 1995 (seen EPA540-S95-504 free downloadable). At the Dutch standardisation institute we have simplified the method somewhat and the new standard will be applicable for all parameters. Metals will keep their 0.45 micron filtration for comparability reasons.

To see how Rhizon soil solution samples work view our video here https://www.vanwalt.com/video-link/rhizon-soil-solution-samplers.html