Calibration: It's what's not in the water that counts
Martin Rowe, Senior Technical Editor -- Test & Measurement World, 5/5/2006 9:38:00 AM
 |
| Figure 1. When properly frozen, triple point of water cells provide a stable reference temperature for calibrating temperature probes. Courtesy of Fluke/Hart Scientific. |
The TPW is the most important temperature for calibrating temperature probes. Defined as 273.16 K (0.01°C), it's one of several temperature points defined in the International Temperature Scale of 1990 (ITS-90) and it's the easiest fixed-point temperature to achieve, or realize (Ref. 1). Other fixed points include the freezing/melting points of metals such as tin, zinc, aluminum, and mercury (Ref. 2).
To realize the TPW, you need extremely pure and consistent water contained in a sealed TPW cell (Figure 1). Unfortunately, not all water is created equal. Although water is always H2O, both hydrogen and oxygen exist in more than one form. Most hydrogen has an atomic mass number of 1, but some hydrogen atoms contain a neutron, which raises the atomic mass to 2. This hydrogen isotope is called deuterium. Oxygen, which mostly has an atomic mass of 16, also exists in O17 and O18 forms with the addition of one or two neutrons (Ref. 3). Water can thus exist with these heavier atoms.
To realize a consistent TPW temperature worldwide, metrology labs need water with a consistent isotopic composition. Fresh water available from rivers and ponds—continental water—varies in composition from source to source. That poses a problem from one manufacturer of TPW cells to another. "We used to use local water in our cells," said Mingjian Zhao, director of primary standards engineering at Fluke-Hart Scientific. "Now we use ocean water with the same isotopic composition as Vienna Standard Mean Ocean Water (VSMOW) to comply with ITS-90." Fortunately, deep ocean water taken offshore has a consistent isotopic composition regardless of which ocean it's taken from. Ocean water eliminates the isotopic variability of continental water.
The differences in isotopic composition of water result in differing TPW temperatures. "TPW variations between VSMOW and continental water can be as much as 70 microKelvin," said Zhao. "That wasn't significant a few years ago," he continued. "Today, temperature-measurement technology has advanced to the point where that makes a difference, so we changed the water in our cells."
The water in a TPW cell must be as pure as possible. The ocean water that arrives at Hart Scientific goes through a triple-distilling process to produce purity better than 99.9999%. The water's isotopic composition could change during manufacturing because of the distillation and degassing process. "We have developed a manufacturing process to make a TPW cell with isotopic composition close to VSMOW," said Zhao.
After purification, the water goes into a borosilicate glass casing under high-vacuum conditions to remove most of the air from the cell. TPW cells were originally made from borosilicate glass only, but now they come in fused-silica quartz glass as well. "Boron from the glass can leach into the water over time," said Fluke-Hart Scientific application specialist Jon Sanders. "The impurities can cause the triple-point temperature to drift from 4 microKelvin to 40 microKelvin. As far as we can tell, no impurities enter the water from the fused silica material."
 |
|
Figure 2. A maintenance bath can keep water in a cell at the triple point for several months. |
REFERENCES
1. Sostmann, Henry E., "A Technical Guide and Standard for the Qualification and use of Water Triple Point Cells," 1999. www.its-90.com/wtpguide.html.
2. "ITS-90 fixed-point cells," Fluke-Hart Scientific, American Fork, UT. www.hartscientific.com/products/fixed-points.htm.
3. "What is the isotopic composition of ordinary water?" International Association for the Properties of Water and Steam, July 19, 2005. www.iapws.org/faq1/isotope.htm.
