VSMOW Triple Point of Water Cells: Borosilicate versus FusedQuartz

M. Zhao1,3 and G. F. Strouse 2

1 2 3

Fluke Corporation, Hart Scientific Division, American Fork, Utah 84003, U.S.A. National Institute of Standards and Technology, Gaithersburg, Maryland 20899, U.S.A. To whom correspondence should be addressed. E-mail: [email protected]

ABSTRACT

To investigate an ideal container material for the triple point of water (TPW) cell, and reduce the influence to the triple-point temperature due to the deviation of the isotopic composition of the water, we developed and tested both borosilicate and fused-quartz glass shelled TPW cells with isotopic composition substantially matching that of Vienna Standard Mean Ocean Water (VSMOW). Through a specially designed manufacturing system, the isotopic composition, D and 18O, of the water in the TPW cell could be controlled within 10 (per mil) and 1.5 respectively, resulting in control of the isotopic temperature correction to better than 8 K. Through an ampoule attached to the cell, the isotopic composition of the water in the cell could be analyzed individually. After manufacture, the initial triple-point temperature of the two types of cell were measured and compared to assess the quality of the cells and manufacturing process. Cells fabricated with the new system agree to within 50 K. Two innovatively-designed borosilicate and fused-quartz TPW cells were made, each with six attached ampoules. We removed one ampoule every six months to track any changes in purity of the water over time.

KEY WORDS: isotopic composition; ITS-90, TPW cell; Vienna standard mean ocean water; VSMOW; water impurities; water triple point.

1. INTRODUCTION

In 1999, National Research Council Canada (NRC) data suggested that the triple-point temperatures of well-behaved triple point of water (TPW) cells decrease approximately 14 K per year, due to impurities leaching from the borosilicate glass envelope into the water [1]. A more recent and intensive study by Ken Hill (NRC) gives a decrease of 4 K per year [2]. The study identified boron, sodium, aluminum, and silicon as the dominant impurities in the TPW cells. These results suggest that the borosilicate glass is a less-than-ideal container, as it is a probable source of an increasing amout of contaminating elements as a function of time. To determine if fused-quartz is a better container material for TPW cells, the National Institute of Standards and Technology (NIST) and Fluke Corporation, Hart Scientific Division (Hart Scientific) began a cooperative study in 2005.

The TPW realization temperature depends on the isotopic composition of the water as well as chemical impurities. The results of an international comparison of TPW cells [Comit Consultatif de Thermomtrie Key Comparison 7, (CCT K7)] for 21 participating national metrology institutes (NMIs) showed a disagreement of about 85 K, with only two NMIs making corrections for isotopic composition [3]. In 2005, the isotopic composition of the TPW was defined to be the isotopic equivalent of Vienna Standard Mean Ocean Water (VSMOW) for the purpose of defining the kelvin [4].

For this study, four TPW cells were fabricated with fused-quartz envelopes and two TPW cells with borosilicate glass envelopes. These TPW cells were sent to NIST for intercomparison. To reduce the uncertainty due to the deviation of the isotopic composition in the cell from VSMOW, we used ocean water as the water source for all TPW cells in this cooperative study. During manufacture of the TPW cells, the isotopic composition of the water was well controlled to substantially match that of VSMOW. The six TPW cells were intercompared against NIST standards every six months from June 2005 to October 2006. Additionally, two innovatively-designed TPW cells (one borosilicate envelope and one fused-

quartz envelope) were made, each with six attached ampoules. One ampoule was removed every six months (method of removal described in section 2.3), providing means to analyze the water periodically and track the purity of the water over time. All ampoules were sent to NIST for purity analysis. The results of the impurity analysis were compared with the intercomparison results in an attempt to quantify the impact of impurities present during manufacture and the effect of the time-elapsed leaching of impurities from the different types of glass over time on the TPW realization temperature. We report these results in a separate paper [5].

2. MANUFACTURE OF TPW CELLS

2.1. Cleaning process

Meticulous cleaning of all parts, including the cell envelope, ampoules, and borosilicate glass tubes in the distillation and vacuum system, is extremely important in the manufacture of TPW cells. All parts are soaked in a proprietary detergent solution for a few days to remove organic impurities, and then flushed with deionized water. The parts are cleaned with a proprietary acid mixture to remove all metallic contaminants, and then flushed with highpurity water (with resistivity higher than 17.8 Mcm). Finally, all parts are steam cleaned for two days. The cleaning procedures for borosilicate glass and fused-quartz are the same, but the acid mixtures are different.

2.2. Water purification and vacuum system

For this study, a borosilicate glass system was designed that allowed for ocean water purification by triple distillation, water filling into the cell, degassing, and sealing of the TPW cells. Through this system, the isotopic composition of each step in the process can be controlled so that the isotopic composition of the water contained in the sealed TPW cell substantially matches that of VSMOW. As shown in the schematic drawing in Fig. 1, this

system is designed so that TPW cells with either borosilicate glass or fused-quartz envelopes can be readily manufactured.

On the vacuum system side of the TPW cell manufacturing system, the vacuum system is composed of a four-inch glass diffusion pump and a mechanical pump. The pumping speed of the system is 300 Ls1, and the ultimate pressure is lower than 1.3 10 5 Pa) (1 107 torr). There are three liquid-nitrogen traps and one water-cooled condenser. The traps prevent oil vapor from the pumps from reaching the TPW cell and the condenser prevents water vapor from the cell from getting into the pumps.

Prior to filling the TPW cell, the ocean water is filtered then purified through a tripledistillation process. To avoid any possible contamination of the pure water, the system was constructed so that there are no valves between the final condenser and the TPW cell, since valves are a potential source of contaminants for the pure liquid water. Once the water condenses after the final distillation stage, it flows directly into the TPW cell through clean glass tubes.

To allow both fused-quartz and borosilicate-glass TPW cells to be produced with this TPW cell manufacturing system, a special glass transition tube with varying thermal expansion coefficient along its length is used to attach a fused-quartz envelope cell.

2.3. Isotopic composition control

Even with VSMOW-equivalent ocean water being used as the water source, the manufacturing process can impact the isotopic composition of the water contained in the manufactured TPW cell (e.g. isotopic fractionation), with a corresponding impact on the cell realization temperature. Table I shows how the isotopic composition and subsequently the TPW realization temperature correction with respect to VSMOW-equivalent water changes during the manufacturing stages. The isotopic compositions of the test cells were tested at the

Stable Isotope Ratio Facility for Environmental Research (SIRFER), University of Utah. All SIRFER measurements are traceable to NIST. Measurement uncertainties (k=1) vary slightly but average about 1 for D, and 0.1 for 18O. Water isotopic compositions are usually reported as per mil () deviations (D and 18O for deuterium and 18O, respectively) from VSMOW.

Filtering the source ocean water to remove particulate matter does not significantly affect the isotopic composition of the water. However, distillation depletes the water of heavy isotopes and degassing enriches the water. The net enrichment or depletion depends on the number of stages and temperature of distillations, and the period and temperature of degassing. It is apparent that even if the source water is VSMOW equivalent, the isotopic composition of water contained in TPW cells can vary if the manufacturing process is not well controlled or understood. Ideally, the isotopic composition of the water contained in every manufactured TPW cell should be individually analyzed. This would allow the user to make a temperature correction based on the unique isotopic composition of the water in their TPW cell.

The water sample used for isotopic analysis should be directly taken from the TPW cell after manufacturing so that the isotopic composition of the water sample is the same as that of the water in the TPW cell. As shown in Fig. 1, the method for doing this is by attaching one or more 20 ml ampoules by fusing an extension attached to the TPW cell. After the TPW cell is manufactured, an ampoule is then filled with about 10 mL of water, flame sealed and separated from the TPW cell, and then sent to SIRFER for analysis.

It is now desirable for NMIs and some primary standard laboratories to analyze the isotopic composition of every newly manufactured TPW cell individually, but this can be difficult and expensive. However, for other applications where such control of uncertainties is less critical, analysis of individual TPW cells is not required. In either case, an uncertainty due to the water isotopic composition, based on either the uncertainty of the isotopic analysis or the uncertainty of the estimate, must be evaluated as an uncertainty to the TPW cell realization

temperature correction. The isotopic temperature correction and corresponding uncertainty of a TPW cell with an unknown isotopic concentration may be estimated using the method described in Ref. 5.

To control the isotopic composition of the water in the TPW cell, it is important to pay attention not only to the isotopic composition of the original water, but also the manufacturing technique. As evidenced in Table I, the TPW cell manufacturing system systematically controls the isotopic composition at each step in the process so that the manufactured TPW cells contain nominally VSMOW-equivalent water. As of January 2007, the TPW cell manufacturing system was used to produce over 200 TPW cells. Before May 2005, isotopic compositions were not very well controlled. After a few months of experience with the new TPW cell manufacturing system, the isotopic composition in terms of D and 18O could be controlled to within a tolerance of 10 and 1.5, respectively, which translates to a TPW cell realization temperature that is within 8 K of VSMOW-equivalent water. As shown in Table II, based on the isotopic composition analysis, the isotopic temperature correction for the 20 TPW cells manufactured from August 2005 to January 2007 is 3 K.

2.4. Special TPW cells for isotopic composition and impurity analysis

To track the rate of impurity leaching from the containment envelope to the water sample and to determine whether fused quartz might be a better material for TPW cells, specially designed TPW cells (one with borosilicate glass and one with fused-quartz envelopes) were manufactured in 2005. Each cell has six attached ampoules. We removed one ampoule every six months, providing means to analyze the water periodically and track the purity of the water over time. Figure 2 shows this uniquely designed TPW cell. After manufacturing, two ampoules from each cell were removed immediately. One was sent to SIRFER for isotopic composition water analysis, and another was sent to NIST for water impurity analysis. Afterwards, we removed one ampoule from each cell every six months and sent it to NIST for water impurity analysis. This allowed us to compare the rate of elemental leaching from the

borosilicate glass and fused-quartz envelopes into the respective water samples. The results of impurity analysis over two years are reported in reference [5].

3. INITIAL TEST RESULTS

After the development of the VSMOW-equivalent TPW cells with borosilicate glass and fused-quartz envelopes, the initial triple-point temperatures of the two types of cells were measured at Hart Scientific to assess the quality of the cells and manufacturing process. The Hart Scientific direct-comparison measurement uncertainties (k=2) are estimated to be 50 K. Eighteen cells were inter-compared at the Hart Scientific primary standard laboratory. Three TPW cells, S/N 1200, S/N 1430, and S/N 1457, were made with the original TPW cell manufacturing system built in 1994. The water source for these three cells was continental surface water. The isotopic compositions were not tested individually, but we sampled the water used to fill the cells. Cell S/N 1200 was made in 2000, and the other two cells were made in 2005. The inter-comparison results are shown in Table III. The results show that the realization temperature differences among the TPW cells made in 2005 are within 50 K, which is within the 50 K expanded uncertainty (k=2) of the comparison tests. However, the TPW realization temperature of S/N 1200 cell made in 2000 differed by 80 K. The drift in realization temperature of the borosilicate TPW cell (S/N 1200) is estimated to be 13 K per year, which agrees to within measurement uncertainties with the results found in Refs. 1 and 5.

4. CONCLUSIONS

A new borosilicate glass TPW cell manufacturing system was developed so that both fused quartz and borosilicate glass TPW cells can be produced with this system. The quality of the cells and manufacturing process were assessed through initial intercomparison testing. The isotopic composition of the water in the cells could be tested individually through a 20 ml ampoule attached on the cell. Using ocean water as the water source, the isotopic composition

of the water in the TPW cells substantially matches that of VSMOW. The isotopic composition, D and 18O, of the water in these manufactured TPW cells can be well controlled within 10 and 1.5 respectively. For TPW cells without an individual isotopic composition analysis, the temperature realization deviation from VSMOW equivalent water is within 8 K. Specially designed TPW cells made from borosilicate glass and fusedquartz were manufactured with six ampoules attached to each cell that allowed the water in the cell to be analyzed periodically and the rate of impurity leaching from the envelope to be evaluated. We report results of the impurity analysis and the intercomparison results performed over two years at NIST in a separate paper [5].

ACKNOWLEDGMENTS

The authors would like to acknowledge the assistance of Hart Scientific Staff: Deming Chen, Rong Ding, and Rick Walker for their considerate advice, and also thank Wayne Newland for his assistance with the illustrations and publishing of this paper. The authors would also like to express their gratitude for the contributions made to this project by many other professionals at NIST and Fluke Corporation, Hart Scientific Division.

DISCLAIMER

Certain commercial equipment, instruments, and materials are identified in this paper in order to adequately specify the experimental procedure. Such identification does not imply recommendation or endorsement by the NIST.

REFERENCES

1. K. D. Hill, in Proc. TEMPMEKO 1999, M. de Groot, ed., (Edauw & Johannissen bV, Netherlands, 1999), pp. 6873. 2. 3. K. D. Hill, Metrologia 38: 7982 (2001). M. Stock et al, Metrologia, Tech. Suppl. 43:, 03001 (2006).

4. Technical annex for the International Temperature Scale of 1990 (ITS90), Consultative Committee for Thermometry (2005); available at http://www.bipm.org/utils/en/pdf/MeP_K_Technical_Annex.pdf. 5. G. F. Strouse and M. Zhao, in Proc. TEMPMEKO 2007, (in press).

Table I. Water isotopic compositions [per mil () deviations from VSMOW] and impact on the TPW realization temperatures incurred during the Hart Scientific TPW cell manufacturing process. Isotopic Ocean After After Distillation Compositions Water Filter First Second Third 2.9 0.7 12 22 31 D () 18 0.3 0.4 1.8 3.6 4.9 O (l) a 1.3 0.3 9.3 16.5 22.9 T, K a The calculation of T is based on the D value only. After Degassing
18 3.2 13.6

Table II. TPW cell water isotopic compositions [per mil () deviations from VSMOW] and corresponding TPW realization temperature corrections for TPW cells manufactured from August 2005 to January 2007. Measurements were performed at Hart Scientific. TPW Cell
G5028 G5035 G5036 G5037 Q1008 Q1012 Q1019 Q1021 Q1022 Q1025
a The

D,
2.0 1.0 1.0 1.0 2.0 1.8 1.0 0.2 0.2 0.7

18O,
0.2 0.5 0.1 0.2 0.3 0.3 0.0 0.3 0.3 0.1

T, K a
2.2 1.5 0.1 0.1 0.6 0.5 1.5 0.9 0.9 1.3

TPW Cell
Q1027 Q1028 Q1030 Q1031 Q1035 Q1039 Q5012 Q5015 Q5016 Q5017

D,
1.0 2.0 2.0 3.0 3.0 1.0 2.0 5.0 1.0 2.0

18O,
0.3 0.2 0.4 0.9 1.0 0.3 0.1 0.7 0.2 0.7

T, K a
1.5 0.6 0.6 2.9 2.9 1.5 0.6 2.8 0.1 2.2

calculation of T is based on the D value only.

Table III.
Cell S/N Q1001 1200 d 1430 d 1457 d G1001 G1003 G5002 G5003 Q1002 Q5004 Q5005 Q5006 Q5007 Q5008 Q5009 Q5010 Q5011
a TPW

Initial inter-comparison results of various TPW cells in 2005.

Container Material Fusedquartz Borosilicate Borosilicate Borosilicate Borosilicate Borosilicate Borosilicate Borosilicate Fusedquartz Fusedquartz Fusedquartz Fusedquartz Fusedquartz Fusedquartz Fusedquartz Fusedquartz Fusedquartz Water Sources Ocean Continental Continental Continental Ocean Ocean Ocean Ocean Ocean Ocean Ocean Ocean Ocean Ocean Ocean Ocean Ocean D () 5.0 99 99 99 10.0 5.0 5.0 0.0 14.0 7.0 0.3 3.0 3.0 2.0 7.0 6.0 7.0 18O () 0.5 12 12 12 1.3 0.8 0.3 0.5 1.9 0.6 0.5 0.5 0.5 0.3 0.6 1.4 1.6 T / K measured a 149 91 100 4 21 6 4 28 27 10 23 13 13 1 7 9 T / K after correction b, c 15 80 22 31 5 15 12 2 16 34 8 23 13 16 8 5 12

Year Made 2005 2000 2005 2005 2005 2005 2005 2005 2005 2005 2005 2005 2005 2005 2005 2005 2005

cell Q1001 is set as the standard; other cells are compared against Q1001. corrected by isotopic composition. The isotopic composition of cell Q1001 is also considered. C The uncertainty (k=2) of the T measurement is 50 K. d The isotopic compositions are estimated.
b Temperature

FIGURE CAPTIONS

Fig. 1. Schematic drawing of the TPW cell manufacturing system.

Fig. 2. A TPW cell for isotopic composition and impurity analysis. The ampoules removed periodically for water testing are shown.

FIG. 1

FIG. 2