News | March 5, 1999

The Importance of System Calibration and Verification for a TOC Instrument

By Karen Clark, John Stillian and Jerry Kirkpatrick, Anatel Corporation

Contents


  • Common Methods For Measuring TOC

  • Application Of TOC Methods

  • On-Line TOC Analysis

  • Calibration Of Conventional Analytical Instrumentation

  • Special Requirements For Calibration Of On-Line Toc Instruments

  • Calibration, Calibration Verification, And System Suitability

  • Conclusion
  • Summary
    Measuring total oxidizable carbon (TOC) in water has become a standard quality test for many industries. However, the accuracy of the measurement is only as good as the calibration of the instrument. This paper discusses the principles of operation for TOC analyzers and outlines several methods appropriate for analyzing differing concentrations of TOC. Issues surrounding measuring and calibrating on-line TOC analyzers will also be discussed. Details and frequency of calibration procedures will be presented.

    Introduction (back to top)
    TOC is the acronym for total oxidizable carbon, often referred to as total organic carbon. Original TOC methods were developed to help correlate information obtained from chemical oxygen demand (COD) and biochemical oxygen demand (BOD) tests in drinking and waste water. The TOC methods were designed to be more automated and efficient than the COD and BOD tests. BOD tests usually take 5 days to complete and COD test involved the use of hazardous chemicals. During the past decade the importance of measuring TOC in drinking water and waste water, as well as the water used in the pharmaceutical, semiconductor and power generation industries, has increased dramatically. Today new government regulations are making TOC analysis a standard test for all types water, in diverse industries.

    The Environmental Protection Agency (EPA) has instituted rules to monitor municipal systems more closely (1). The main reason for measuring TOC in drinking water is that chlorine, the primary drinking water disinfectant, has the potential to react with some organic compounds in the water to form chlorinated hydrocarbons which have been associated with carcinogenic activity (2). Drinking water is the initial ingredient for most water purification systems. Monitoring this feedwater is critical to maintaining the overall efficiency of any water purification process. The water's characteristics such as hardness, particulates, bacterial levels, conductivity, and TOC levels significantly influence downstream processing. It is important to remember that seasonal changes can also effect the water's composition.

    TOC content is a common measurement in industrial wastewater, as well as general industrial discharges. It is important to measure the TOC levels in effluent wastewater to determine if the water can be reused or reclaimed for additional processing. This is the first step in effective waste-minimization and reduces any unnecessary harm to the environment.

    The power generation industry recognizes TOC as a significant corrosion contributor. TOC contamination found in the water will become corrosive acids at higher pressures, which can reduce the life of boilers, reactors and turbine blades. High-purity water is necessary for continuous operation of power facilities.

    TOC levels in ultrapure water such as that used in the semiconductor industry has been found to be a major contributor to increased product defects (3). Today, device geometry reductions accompanied by increases in circuit densities are imposing challenging demands on the purity of the process water. Semiconductor manufacturers must monitor TOC levels at all stages of water purification, most importantly at the point-of-use. Since even the slightest change in TOC levels can affect production yield, the industry has incorporated on-line TOC monitoring to provide trend information throughout the entire water purification process (4). Today, TOC levels in semiconductor process water are being controlled to below 1 ppb.

    Finally, and most recently, TOC test methods are now a part of the pharmaceutical industry. The United States Pharmacopeia (USP) has made TOC analysis a standard quality test for production of Purified Water (PW) and Water For Injection (WFI) (5). Measurement of TOC is a direct reflection on the quality of the water being produced and can have a significant impact on the manufacturing process of pharmaceutical and bio-pharmaceutical products.

    CommonMethods For Measuring TOC (back to top)
    TOC measuring devices share the basic technology of converting all the carbonaceous material in a given sample to a form that is more readily measured, such as CO2. The oxidation and detection technologies vary based on the specific analysis method, and each technique is appropriately applied to a different concentration of TOC.

    Combustion Oxidation: The combustion method measures total carbon (TC). It requires samples injection by syringe into a high-temperature furnace with a platinum or cobalt catalyst. This process oxidizes all of the carbon materials present to CO2. The CO2 is swept into a non-dispersive infrared (NDIR) detector by a carrier gas usually nitrogen for final measurement. The amount of CO2 measured is directly proportional to the amount organics present in the original sample.

    A variation of this method employs a stream splitter which directs equal parts of the sample to two furnaces at different temperatures. One furnace at 150º C measures the total inorganic present and the other at 950º C measures the TC present. The final TOC amount is then calculated by:

    TOC = TC - TIC

    Flame Ionization Detection: This method involves reducing CO2 that is produced to methane that can be measured by a flame ionization detector. The methane is mixed with either nitrogen or helium carrier gas and is carried to the detector where it is burned in a hydrogen/air flame. The quantity of ions produced upon burning the methane is subsequently measured and used to calculate TOC content .

    Wet Oxidation: Wet oxidation involves adding an acid to the water sample to reduce the pH to approximately 2 to 3. At this low pH any inorganic carbon that is present is liberated as CO2 into a nitrogen carrier gas and is directly measured by an NDIR detector. Any remaining carbon in the sample is assumed to be TOC. A persulfate oxidant is added to the sample, and in the presence of heat and/or UV radiation, the remaining carbon is oxidized to CO2. The amount of CO2 generated is then measured by the NDIR to determine the amount of TOC.

    Colorimetric: Similar to wet oxidation, the colorimetric method involves acidifying and sparging the water sample to release any inorganic carbon that is present. A chemical oxidant is added to water sample and along with UV radiation, the organic carbon is oxidized to CO2. The CO2 that is produced is passed through a semipermeable membrane and dissolved into a dilute buffered phenolphthalein solution. The solution changes color based on the increasing pH caused by the dissolved CO2. The color change is measured spectrophotometrically at 530 nm. The intensity of the color change is directly proportional to the amount of TOC present.

    Conductivity: Conductivity based TOC methods oxidize the TOC that is present to CO2 using UV radiation typically in the presence of a titanium oxide catalyst. The CO2 that is produced re-dissolves into water as bicarbonate (HCO3-) and hydrogen ions (H+). These ions increase the conductivity of the water. A conductivity meter or probe is then used to measure the change in conductivity of the solution. The change in conductivity is proportional to the amount of TOC present in the water.

    Application Of TOC Methods(back to top)
    As mentioned previously, the measurement of TOC in water is important to many different industries. The technology chosen to analyze for TOC will depend on a number of variables. The most important variable to consider is the water source. Is the water to be measured wastewater, raw water, drinking water, high purity water or ultrapure water? The water source to be measured is a major factor in considering which TOC method to use.

    Typically, wastewater and raw water contain both soluble and insoluble organic material in high concentrations (> 100 ppm). Since Combustion Oxidation oxidizes all of carbon materials present by the use of a high temperature furnace, this method is suitable for measuring insoluble organics found in wastewater and raw water. Drinking water typically contains soluble TOC at levels less than 10 ppm. The TOC in this water can easily be measured by Wet Oxidation. Measurement of TOC in high purity water and ultrapure water used in the Pharmaceutical and Semiconductor industries respectively, requires a more sensitive method like the Conductivity based method. TOC levels in these waters range from less than 500 ppb for pharmaceutical high purity water to less than 1 ppb for semiconductor ultrapure water.

    Instruments for TOC analysis are of two types, laboratory based or on-line. Laboratory instruments require that "grab" samples be collected from the particular water source or purification point and brought to the laboratory for analysis. On-line analyzers have been developed for use directly at the process point to provide immediate control information. Laboratory analyzers are well suited for the measurement of high TOC waters or for measuring water where a small amount of contamination will not affect the overall result or where loss of volatile organics is not a concern.

    It has been well documented that laboratory TOC analysis requires special handling of samples including meticulously cleaning all glassware used to collect samples and prepare standards (6). Despite proper procedures, it is not unusual to see a significant amount (a few hundred ppb or more) of TOC contamination not associated with the original water sample. In high TOC waters this extra contamination may be negligible, but in high purity water or ultrapure water where the starting TOC levels are less than 500 ppb, this level of extraneous contamination can have a detrimental impact on reporting results. Over the years, the Semiconductor and Pharmaceutical industries have relied on on-line TOC analyzers for TOC testing. The major advantage of on-line analyzers for these industries is the ability to continuously measure their water purification at points where TOC excursions can cause a decrease in water quality, final product quality and ultimately production yields.

    On-Line TOC Analysis (back to top)
    Virtually every component in a high purity water system is a suspected source of organic contamination. Organic fouling can occur in exchange beds or membranes and pass through boilers and stills, resulting in contamination carryover to product water. Organic compounds can degrade or break down to organic acids that can cause corrosion in the piping and complementary vessels. Slight increases in TOC as detected by on-line analysis can provide advance warning of unit operation failure including carbon bed exhaustion, membrane breakdown or sloughage from resin beds.

    In high purity water applications, on-line TOC monitoring leads to better process control and can be used to develop and maintain standard operating procedures as well as validation of the entire water system (7). However, on-line TOC instruments present a unique set of issues related to calibration and operational verification particularly in regards to the Pharmaceutical industry where conventional analytical instrumentation is also used.

    Calibration Of Conventional Analytical Instrumentation (back to top)
    Calibration is defined as checking physical measurements against accepted standards (8). Typically, a response profile of an instrument is generated from standards over a given concentration range that unknown samples will be analyzed. The data that is generated is usually plotted as a calibration curve.

    Calibration curves approximate a straight line with defining constants including the slope (m), y-intercept (b) and correlation coefficient (R). The correlation coefficient is a statistical parameter that is a determinant of the linear relationship between the analyte concentration and the measured response. A correlation coefficient value of ³ 0.990 is typically representative of a good linear relationship. These constants are estimates of the true values and are considered performance characteristics of the instrument. Measuring replicates of the calibration standards can minimize the deviation of the data from the line. The magnitude of such deviation is a function of several variables including precision, accuracy, number of standards used and concentration of analytes.

    Why is calibration important for TOC analysis? With any analytical method or technique, the required information is obtained by measuring a physical property that is characteristically related to the component of interest. Water has certain physical properties that can be measured directly, for example pH and conductivity. The amount of total organic carbon present in a given water source can not be measured directly. In all cases the organic carbon that is present must be converted in some manner to CO2 before the actual concentration present can be determined. Because TOC is converted to another entity, calibration is the way to determine if the analytical method of choice is functioning properly. In other words, TOC is being properly converted to CO2 and the amount of CO2 produced is being measured accurately.

    Special Requirements For Calibration Of On-Line Toc Instruments (back to top)
    One of the most important issues to consider about calibrating on-line TOC instrumentation is introduction of multiple concentrations of standards. During normal operation a sample stream is continuously supplied to the instrument. During calibration there must be a bypass stream available for standard introduction. Calibration standards must be introduced in a similar manner as the sample stream to prevent any biases. Sample introduction must also be convenient so that the instrument does not have to be disconnected from the process line for calibration. An example of such an instrument is shown in Figures 1 and 2.

    Sample analysis mode of an on-line TOC analyzer

    Calibration Mode of an on-line TOC analyzer

    The method of analysis for this on-line TOC analyzer is based on a single platform for both conductivity measurement and TOC analysis. Another issue to think about in regards to calibration is how are the standards themselves prepared. Are they prepared and packaged to prevent extraneous contamination as well as maintaining stability of the TOC standard?

    Calibration, Calibration Verification, And System Suitability (back to top)
    Calibration is a rigorous test of an instrument's performance, and defines the instrument's response versus concentration for subsequent sample analysis. To insure valid results, multiple test points or standards are used. Calibration is to be performed at set intervals, typically once or twice a year. Calibration Verification, sometimes called validation, is a check of the "validity" of an instrument's calibration. Calibration Verification is usually a single point analysis and is performed more frequently than a full-scale calibration.

    What is System Suitability? System Suitability is an industry specific test protocol and intended to be a substitute for instrument Calibration. System Suitability is defined as the process of validating whether your instrument is acceptable for providing useful analytical data without bias and is appropriate for the intended application (8). This is typically done by analyzing a material that easily provides a response and compares it to the response of a material that is hard to detect, or in the case of TOC, hard to oxidize. All analytical methods, not just those used for TOC analysis, should include a test for the measurement of the system's suitability. An example of how system suitability can be used for TOC analysis can be found in the pharmaceutical industry. The USP TOC method <643> defines how system suitability is to be performed on a TOC instrument (10). Initial requirements for TOC instrumentation include detection limits of at least 50 ppb and that any instrument used for TOC analysis must be previously calibrated. This TOC method is specific for the determination of TOC levels in Purified Water and Water for Injection.

    Anatel's A-2000 TOC Analyzer

    The first part of system suitability involves measuring the instrument's response for three solutions, sucrose (500 ppb as carbon), 1,4-benzoquinone (500 ppb as carbon) and the water used to prepare or reconstitute the standards. The sucrose and 1,4-benzoquinone must be prepared from USP reference standards.

    The instrument's response efficiency is then calculated as follows:

    Response Efficiency = [(Rss - Rw)/(Rs - Rw)] x 100

    where:

    Rss = instrument response for 1,4-benzoquinone (500 ppb as carbon)
    Rs = instrument response for sucrose (500 ppb as carbon)
    Rw = instrument response for reagent water

    In order for an instrument to be "suitable" for measuring 500 ppb levels of TOC in purified water and water for injection, the response efficiency must be no less than 85% and no greater than 115%. The USP method states that all water tested must have a TOC value of less than or equal to the limit response (Rs - Rw) in order to be considered acceptable.

    One question that is always asked is how often should Calibration and System Suitability be performed? This is a very difficult question to answer consistently because it depends on a number of factors. First of all, test frequency is water system dependent. Some water systems are very stable and do not change very much; others have frequent fluctuations in water quality. Second, test frequency is instrument dependent.

    If your instrument is reliable and stable, frequency of the tests will be minimal. If the TOC instrument is more complicated than some on-line analyzers such that carrier gases and chemical oxidants are used for analysis, Calibration and System Suitability should be performed more often. A typical test to determine a need for a new calibration is to run a Calibration Verification.

    If the single point analysis of a Calibration Verification is out of specification then chances are the instrument needs to be re-calibrated. A good analytical "rule of thumb" is to run a system suitability every time an instrument is re-calibrated. Also, the frequency of system suitability should be determined systematically from test results. For example, run a System Suitability once a week for four weeks. If the tests fall within specifications run a System Suitability once a month for six months. If the tests still fall within specifications then it is likely that the system is stable for six months and that System Suitability only needs to be performed at that interval. Always make sure historical data is collected to show stability of the water's TOC analysis system and validity of the chosen System Suitability interval.

    Conclusion (back to top)
    TOC has become and important marker for water quality in many industries. There are different methods available for measuring TOC, but it is important to consider what the water source is and the physical attributes of that water as well as what the expected TOC levels will be, in order to determine which technology is appropriate for accurate TOC analysis. For high purity water and ultrapure water applications additional consideration needs to be made in regards to laboratory TOC analysis or on-line TOC analysis. On-line TOC analysis leads to better process control and provides advance warnings of water system upsets.

    Since TOC is not an absolute measurable parameter like pH or conductivity, the performance of TOC instruments needs to be verified by standard analytical techniques including calibration, calibration verification, and system suitability. These techniques involve measuring standards that need to be handled in the same way as the sample stream to prevent any bias in the measurements. Finally, the frequency of the performance tests will mainly depend on the individual water system, but actual intervals should be determined systematically.

    References

    1. Environmental Protection Agency Information Collection Rule 61-FR 24354-34388 (May 1996).
    2. W.S. Joyce et al., "THM Precursors in the Environment", J. Amer. Water Works Assoc. 76 (6), pp. 102-106 (June 1984)
    3. P. McConnelle et al., "Water Quality Improvements and VLSI Defect Density", Semiconductor International (August 1996).
    4. Poirier S. and R. McIntosh, "The Application of On-line TOC monitoring in Semiconductor and Power Generation Plants", Ultrapure Water (July/August 1995).
    5. United States Pharmacopeia (USP) 23, Fifth

    Supplement (November 1996).

    6. M. Datta and A.K. Vickers, "A Quantitative Study of the Effect that Various Sampling techniques have on Total Organic Carbon Analysis", Pittsburgh Conference on

    Analytical and Applied Spectroscopy, New Orleans, LA (March 9-13, 1992).

    7. Hjelle, M.A., M. Bauer, N. Cohen, "Characteristics of an Ultrafiltered Recirculating Deionized Water System Used in the Manufacturing of Implantable Biomedical Devices', Ultrapure Water 10 (8), pp. 31-35 (Nov 1993).

    8. Taylor, J.K "Validation of Analytical Methods", Anal. Chem. 55, 600A-608A, (1983).

    For more information: Karen Clark, Anatel Corp., 2200 Central Ave., Boulder, CO 80301. Tel: 303-442-5533. Fax: 303-447-8365.