For experiments on geological samples in a closed (hermetically sealed) environment it is necessary to know their moisture content, because the material properties are very sensitive to changes in this quantity. I expect that it will have to be calculated from the humidity of the air inside the container which depends on the water content of the sample. A robust and reproducible way of measuring the air humidity to an accuracy better than 1% of the value of the relative humidity is therefore needed. The total air volume will be in the order of magnitude of 1 litre. The usual hygrometers are not accurate enough. Which method(s) could you recommend?
PS (09 Jan 2014): Not only the water content of the sample is of interest, but also whether the adsorption of the water changes during the experiment (we are interested in the capillary effects in the pores of the sample).
Jan-Martin, considering all aspects of your application (sealed system, high humidity and high temperature) I would strongly suggest a chilled-mirror dew-point hygrometer. There are few models equipped with a small head (20-mm OD) with gasket sealing able to work up to 200 bar and to 90 °C. You can screw it in the cap of your cell and preserve the internal environment during the experiments.
I think the better way is to calibrate samples by using a climatic chambers. Secondly, you have sorption curve of your materials and you can weight samples to know relative humidity inside materials. That's what I do for porous building materials.
Good luck for your experiments,
You could use spectroscopic instrumentation to measure the water concentraiotn. With such a method you could reach ppmV.
Good precision! Onset have many others that may satisfy your problems!
http://www.onsetcomp.com/products/data-loggers/u23-001
Your question depends on the method you measure the relative humidity (rH). Most accuracy can be obtained, if you use a high precision dew point meter (uncertainty with k=2 approximately 0.04 °C) and a precision PRT for the air temperature (uncertainty with k=2 approx. 0.02 °C). The resulting uncertainty for the rH taking into account the uncertainty of water vapour equation and enhancement factor is approximately in the range of 0.25 % in the lower region up to 0.4 % in the higher region.
If you use a good capacitive polymer sensor for relative humidity, you can achieve a calibration uncertainty for the moment of calibration which is only slightly lower (0.3 % to 0.5 % with k=2). But those sensors often exhibit a drift over time depending on their operating conditions. I would prefer to calibrate polymeric sensors with a reference traceable to National Standards two times a year to ensure continued precision.
Ironically, air humidity is very tricky to measure very accurately. Most relative humidity sensors use some kind of capacity sensor and they are usually not accurate enough (typically to within a few to several % depending on RH and T).
Best relative humidity sensors are probably a proper dew-point hygrometer where a mirror is piezo-electrically heated/cooled and dew on the mirror is detected; the temperature of the mirror at dew formation is the dew point. Expensive but reliable and absolute accurate. Other expensive methods include spectroscopic (e.g. Lyman-a); fast but problems with absolute calibration.
Simplest and hard-to-beat is the simple wet-and-dry termometer; the tepmerature difference between a dry thermometer and one which is constantly wet (and hence evaporates) is a relible, simple and cheap method, although a bit messy.
From your description it sounds like what you want is not relative humidity at all, but absolute humidity.
Thank you for all your input so far! I forgot to mention that the experiments are to be conducted at fixed temperatures, and that the small volume will not be favourable to using bulkier instruments, and will not allow for much air circulation. Whether absolute or relative humidity is measured, does not matter so much, as long as a reliable value for the moisture content of the air can be obtained. The duration of measurements is of no great concern, and saturated salt solutions may be used to provide standards (or reference points) for air humidity. Could spectroscopic methods be used in such a case?
The accuracy depends on your requirements. For the US Climate Reference Network our requirement is +/- 3% over the range from 10-90% RH and +/- 5% for RH values below 10% and above 90%.
The minimum resolution requirement we have is 1%.
As noted in another response, air humidity is very tricky to measure very accurately and so our requirements are in line with about the best we can expect from the instrumentation available.
Jaan-Martin what is you measurand? Are you interestend in measuring air humidity or to infer the moisure content by means of "water activity" measurements?
Precision measurement of air humidity and moisture in material demands quite different approaches. The most accurate way to measure air humidity is via dew point and air temperature measurements, the former being measured by a chilled-mirror hygrometer. As Norbert pointed out you can achieve uncertainty of 0.2-0.3% RH.
Water activity can also be measured in the same way or with polymeric sensor but with a larger uncertainty.
On the other hand, reference methods for moisture content in material are the thermogravimetric analysis and the Karl Fischer titration. However it is difficult to achieve uncertainty below 1% with such methods.
I advise to contact Deutscher Wetterdienst (DWD) - Observatory Lindenberg, Director: [email protected]
They have developed the German national standard method for high precision rH measurements.
Harald, PTB developed and maintains the German national standard for humidity!!
Vito, as far as I remember this was a collaboration with PTB, and Lindenberg still holds the PTB standard.
This is not likely to be a practical route to measuring RH for you, regardless a trapped aerosol particle can be used as a probe of RH with an accuracy better than 0.09%, using cavity enhanced Raman spectroscopy and optical tweezers. As I said this probably doesn't help you but it does show the RH can be measured very accurately by indirect methods.
http://pubs.acs.org/doi/abs/10.1021/jp107802y
Most of things has already been set in the previous posts. I can only confirm on the basis of my own experience that most sensors do not work well above 90% and that indeed an accuracy below 1% is not an easy job. The only devices I really trust are sensors based on dew-point mirrors.
Vito and all, thanks again! The idea was to find out about the water activity (which may differ from the actual water content) in the porous rock sample by measuring the water content of the air above it once an equilibrium has been reached. As I suspected, and as your answers have confirmed, the latter will be difficult to measure accurately, in particular as the available space will be restricted. May I enquire again as to the accuracy and reproducibility of spectroscopic methods?
if you are looking for a very small Chilled mirror, that would fit into your 1 l volume, check METEOLABOR, the Schneewittchen rawin sonde instrument, probably a good solution.
Spectroscopic methods (TDLAS, CRDS or the like) offer very high detection capability (with a sensitivity of 1 ppb or better) but generally have a limited measurement range (maximum few thousands of ppm), which I guess is well below your needed measurement range, and have accuracies not better than 1-2 %
Jan-Martin,
your approach to find moisture content from relative humidity of surrounding air in a sealed container is fundamentally flawed (as you can read from Vito's and Laurent's statements). Of course, if your sample contains enough water it will always saturate the air to the dew point. If the water is enclosed in porous material and the pores are small it will take a long time to establish equilibrium of water exchange and thus a constant value of relative humitidy outside. If water is strongly adsorbed to internal surfaces of the material it may fail to evaporate at all.
Following Ulrich's remarks, I would suggest to look at and compare 3 methods:
1. Evolved gas analysis, where the sealed cell is put in a oven, the seal is broken, a temperature ramp is applied to the sample and the evaporated water vapor detected by MS or coulometric detection. Different levels of bound water can be detected un this way.
2. Similar approach by using oven-coupled coulometric Karl Fischer.
3. A Microwave absorption method which is non invasive and allows to analyse the sample while sealed.
A running EU project covers parts of this aspects and related calibration issues.
Pls. see www.metef.net
My original posting was perhaps formulated inaccurately regarding the reasons for the enquiry. Of course, the total water content of the sample can be found by other methods, but more specifically, we meant to track how the adsorption and desorption of the water in the rock sample, or the suction, changes during the experiment, and expect that these changes will be reflected in the air humidity. Hence the need for an accurate measurement of the latter.
My intention was to indicate that the 'reflection' mentioned in
is not a natural one and will depend critically on non-probe-specific paramters such as volume and agitation of the surrounding air. Not knowing details of your planned experiments I'm certainly unable to see whether this will hurt your results. But, hear my whistle! Nevertheless, good luck!
If I did understand your question properly you want the absolute concentration of water, more closely its time variation with high precision. One spectroscopic method to obtaion this parameter is to measure the OH resonance fluorescence excited by FUV photolysis (Lyman-a or Krypton, 121.7 or 123.7 nm, respectively). The technique may need a complementary method for absolute calibration, though. Good luck !
I think that the method you'll use depends on the RH range you'll be working on.
If you want to detect really low humidity (
We are using http://www.edgetech.com/moisture-humidity/gallery/item/dewmaster-chilled-mirror-hygrometer - you can find the brochure with specifications there. Very good instrument, we use it in the HTDMA setup but be prepared to spend few thousands euro. These instruments are very precise and VERY expensive.
German, this is capacitive sensor. They are not as accurate as the author wishes for.
Panagiotis, we are looking to measure chiefly in the upper range of relative humidity between maybe 60 % and 100 % r. h., and between 20°C and 90°C. It may not be practicable to remove the sensor for calibration purposes as we would like to keep the system closed. The air volume needs to be rather small due to the limited sample size, and the temperature is to be kept constant. It may, however, be possible to force some circulation if the method requires this.
Jan-Martin, considering all aspects of your application (sealed system, high humidity and high temperature) I would strongly suggest a chilled-mirror dew-point hygrometer. There are few models equipped with a small head (20-mm OD) with gasket sealing able to work up to 200 bar and to 90 °C. You can screw it in the cap of your cell and preserve the internal environment during the experiments.
Just to support a comment made some time ago by Mutze, if this is a closed system, it will eventually reach equilibrium, which thus constitutes 100% "relative humidity." If you have free water present and there is circulation of the air, that equilibrium will obtain fairly rapidly and the final water vapor pressure will be determined by the temperature alone (and is easily calculated). If you then replace this atmosphere with dry air, it will again achieve the same water vapor pressure, as long as T is the same and there remains excess water. After the excess water is gone, the equilibrium VP point will drop, consistent with the nature of the adsorption of the remaining water to the solids (the adsorption isotherm); and each subsequent replacement of the atmosphere with dry air will result in further lowering of the water VP. In each case, if you monitor the VP as a function of time, you can get useful information about the rate of desorption of the water. These kind of adsorption processes are usually understood most simply in terms of the BET isotherm.
As some others have noted, spectrophotometric measurements (like IR absorption) can be easily calibrated and can yield very high accuracy and precision, though, depending on the nature of your instrument and light source, such calibration curves may necessarily be nonlinear. But this need not be much of a problem.
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences