Key Comparisons in Electrochemistry Significance to Metrology and Global Trade Kenneth W. Pratt* and David L. Duewer National Institute of Standards and Technology (NIST) 100 Bureau Dr., Stop 8391 Gaithersburg, MD USA NCSLI conference, San Antonio, TX USA 27 July 2009, Session 2D

2 What will I learn from this talk? Key Comparisons (KCs) –where KCs fit in the hierarchy of international metrology –how KCs relate to Certified Reference Materials (CRMs) –how KCs and Pilot Studies differ Lessons learned from KCs –sample integrity is critical –state of the art of the primary measurement of pH can be improved –one must specify the desired measurand exactly How KCs relate to real-world measurements –example from pH metrology

3 The Key Comparison

4 International Metrological Hierarchy Consultative Committees (CC) The CIPM also supervises the International Bureau of Weights and Measures (BIPM).

5 Working Groups of the CCQM 7 Working Groups (WG)

6 Key Comparison (KC) International comparison to assess measurement capability Participants: National Metrological Institutes (NMIs) –use their standard measurement procedure for given measurand WG(s) propose, CC approves –electrochemistry: EAWG proposes, CCQM approves Agreed protocol, time schedule, sample –proposals distributed to WG members and agreed to at WG meeting –must use “uncharacterized sample” One participant is Coordinating Laboratory –verifies sample homogeneity –distributes samples –compiles NMI results (value and uncertainty) –presents report to WG WG agrees on and forwards Final Report to CC for approval

7 How is a KC result used? Assess performance of NMIs –Final results publicly available at NMIs are identified –Electrochemistry: primary measurements highest-level link in traceability to the SI Support Calibration and Measurement Capabilities (CMCs) –CMCs available at –pH and electrolytic conductivity CMCs are under “Chemistry (QM)” Assist in evaluation of candidate CMCs –1 st review: Regional Metrology Organizations –2 nd review: KCWG –Acceptance based on KC results + other available information

8 Participants –NMIs –Designated Institutes (by NMIs) On an international level, used –to establish measurement parameters –as a training exercise –as a preliminary “test” before a subsequent KC Pilot Study results –not binding on participants as are KC results –published only by agreement of all participants –not listed in Appendix B as support for CMCs Otherwise similar to a Key Comparison –often a KC and Pilot are run in parallel Pilot Study International comparison for investigational or training purposes

9 EAWG KCs and Pilot Studies, a Joint with IAWG

10 KC Reference Value (KCRV) Benchmark value for a KC –agreed upon following open discussion by participants –usually obtained by an estimate of central tendency of NMI results weighted mean, median, median average deviation, etc. an independent value (e.g. solution preparation) may be used Deviation from KCRV = Degrees of Equivalence, D i x i = result of NMI i x KCRV = KCRV Expanded uncertainty of D i u = standard uncertainty of quantity in ( ) R = correlation factor between x i and x KCRV

11 KCRV - a Valid Gauge of CRM Uncertainty? If –a 1:1 correspondence exists between a set of KC(s) and CRMs Then –The D i value for an NMI reflects the uncertainty that the NMI can attain in those CRMs pH CRMs: fulfill this requirement closely Conductivity and titrimetric CRMs: not as closely

12 Correspondence between pH KCs and NIST pH Standard Reference Materials (SRM ® ) 1:1 correspondence → KCs reflect SRM performance closely

13 Correspondence between Electrolytic Conductivity KCs and NIST SRMs Lack of 1:1 correspondence between KCs and most SRMs

14 Lessons Learned

15 1. Sample integrity is critical Problem –K9, K17, and P22: bottles leaked during shipment pressure in cargo hold is ≈75 kPa (0.75 atm) –Security issues for carriers –Leakage may cause change in sample composition –Cost of reshipment Solution –Double-bag bottles –Weigh each bottle at Coordinating Laboratory and by recipient agreement should be within 0.01 % –Ship second set of bottles if mass check fails

16 2. pH measurements can be improved Problem –NIST primary pH measurement uncertainty was larger than other NMIs originally discovered in CCQM-P37 Pilot Study problem traced to sluggish response of Ag|AgCl electrode in primary cell –Would not have been discovered except for international comparison Solution –Equilibrate Ag|AgCl electrode in solution to be used next day electrode equilibrates during overnight storage –Uncertainty decreased by factor of 3 after adopting this protocol Uncertainty of NIST pH SRMs was improved as a result

17 3. Specify your measurand precisely “NMI X” needed to verify its standardization of dilute HCl –NMI X: “The concentration is nominally 0.1 mol/kg.” –NMI X sent sample to (experienced) “NMI Y”: NMI X and NMI Y differed –NMI X sent 2 bottles to NIST to resolve the discrepancy. Bottle 1 leaked (see Lesson #1). Bottle 2 was OK. NIST performed coulometric titration of NMI X acid from Bottle 2 NIST reported result as molality and amount content –molality: mol HCl per kg solvent (water) –amount content: mol HCl per kg solution –concentration: mol HCl per L solution –NIST amount content agreed with NMI Y, molality agreed with NMI X Resolution –The discrepancy was in the measurand, not the titration –Neither measurand was the concentration! –Watch your terminology: be specific!

18 How KCs Relate to Practical Measurements

19 Pyramid of Traceability for pH Step 1 Step 2 Step 3

20 Now you know... where KCs fit in the hierarchy of international metrology how KCs relate to Certified Reference Materials (CRMs) how KCs and Pilot Studies differ that sample integrity is critical how the primary pH measurement of pH was improved that one must specify the desired measurand exactly how KCs relate to real-world pH measurements

21 Thank You for Your Attention