Handling samples for 18O-isotope analysis

Ole Humlum, The University Courses on Svalbard (UNIS), Norway.

General

The link between long-term changes in the isotopic composition of precipitation and surface air temperature at a given location is probably the most important relationship as far as paleoclimatic applications are concerned. Today, a semi-empirical  temperature/stable isotope relationship was established for coastal stations in the mid and high northern latitudes. Isotope analyses have been applied for several years within glaciology, but this technique is still relatively new in periglacial- and permafrost-related research.

As this technique is likely to see widespread use in periglacial/permafrost research in the years to come, a short description of how to handle ice or precipitation samples before delivery to a laboratory is given below. For a short background to the isotopic composition of precipitation, click here. 

Sample size

Normal is 20 ml, 5 ml is used as a minimum.

Bottle. 

Use small plastic bottle than can tolerate freezing without rupture. Cap opening should not be smaller that 15 mm.

Storage

Stored non frozen for a maximum of 2 months. Maximum time depends on how clean the sample is. Store in darkness. For prolonged storage, freeze.

Sample treatment

During sample preparation, water and CO2 has to equilibrate. For this process to run, the pH should be lower than 5. For absolutely clean water (inland ice), the CO2 in itself will lower the pH sufficiently. For less clean samples, it may be required to add acid. If acid should be added, this must be done prior to delivering the samples to the laboratory. Use concentrated sulphuric acid. Dip a thin glass pin in the acid, and then in the sample. Avoid evaporation, and be sure that the pH don't get lower than 3,5.

Number of samples

Investigate how many samples the laboratory takes in one batch in order to make efficient use of their working time. Often a standard batch is 115. If less samples are delivered, the laboratory will usually fill up with dummy samples or duplicates.

Packing

Pack the samples, so that consecutive samples are nicely in order. Place a description of the numbering scheme in the box, so its possible to measure the samples in correct sequence.

Standards

Usually the laboratory would appreciate to know the approximate delta value in order to select the working standard. At many laboratories they will then use two different standards for the measurements, one that should be close to the mean value of the samples, and one that is somewhat (15 permille) off. Both sets will be used in the automatic correction of the measurements.

Latest update: 19. December 2002.