by: Philip L. Coduti, Richard L. Hoch, and P. Lawrence Meschi
Pages: 53 - 61; January, 1995

Not only are the quantities of emerging precision cleaning agents, equipment, and processes staggering, but assessing their level of performance relative to previous approaches presents even greater challenges. As environmentally harmful cleaning solvents are phased out and replaced, the quest for implementing alternative technologies raises numerous questions:

• What alternative cleaning agents are best suited for my products?

• How can I modify my cleaning system to achieve optimal results?

• How can I evaluate the effectiveness of one cleaner

versus another?

• How does one cleaning process compare to another?

• How can I verify that I have achieved the level of cleanliness required, or that which is at least equivalent to what I previously had?

Even though the complex variety of cleaning variables raises numerous questions, one underlying commonality addresses them all: How do they ultimately affect the cleanliness levels of my products?

Carbon’s the Culprit

Generally, the most common contaminants present on the surfaces of most materials are carbonaceous in nature. Most are volatile organic carbon (VOC), non-volatile organic residues (NVOR), or both, originating from contaminated cleaner baths, cutting oils, drawing compounds, fingerprint oils, hydraulic fluids, lubricants, machining fluids, rolling oils, rust preventative oils, or solder fluxes and other organic substances introduced intentionally or inadvertently during the manufacturing process.

Heat produced during some manufacturing processes may cause some organic residues to become tenaciously bonded to the surface. Other manufacturing processes can incorporate high-temperature heat treatments during which surface organic residues may char and form inorganic surface carbon (amorphous or elemental carbon residues).

The detrimental effects of surface carbonaceous residues on surface wettability,¹ the corrosion resistance after painting,^2,3 adhesive bonding,⁴ electroplating,⁵ and reactivity with liquid oxygen⁶ are well known.

Cleanliness Verification

Cleanliness can be verified by either indirect or direct methodologies.

Some indirect methods use solvents to extract NVOR from samples or cleaned parts. The amount of organic residue collected in the solvent can be determined by gravimetric or spectroscopic techniques. Indeed, some of the extraction solvents that these indirect methods employ are environmentally sensitive and may become unavailable.

Other indirect methods verify cleanliness by relying on total organic carbon analysis of rinse water used during the last stage of a cleaning process.

Direct cleanliness verification methods incorporate sophisticated surface analytical techniques such as Auger Electron Spectroscopy (AES), X-ray Photoelectron Spectroscopy (XPS), Secondary Ion Mass Spectrometry (SIMS), and Fourier Transform Infrared (FT-IR) spectroscopy.

Although these techniques are very sensitive to surface contaminants, they are costly, can be sensitive to surface texture, and require skilled operators. However, they are very useful for corroborating and/or complementing other direct or indirect cleanliness verification methods.

Another direct cleanliness verification method, known as Direct Oxidation Carbon Coulometry (DOCC), is quantitative, cost-effective, rapid, easy-to-perform, surface-texture independent, and adaptable to production environments. The technique employs in-situ direct oxidation of surface carbon to carbon dioxide (CO₂), followed by automatic CO₂ coulometric detection.

Principles of Operation

During DOCC cleanliness verification analyses, the amount of surface carbon on non-combustible materials like most metals, glasses, and ceramics is converted to CO₂ by combustion with oxygen gas.

Differentiation between organic and inorganic surface carbon is controlled by combustion temperature. The amount of CO₂ produced, proportional to the amount of surface carbon present on the sample, is detected and measured by automatic CO₂ coulometric titrimetry.

Since coulometric detection of CO₂ is an absolute measurement based on Faraday’s Law, the preparation of chemical standards for calibration is not required. This useful aspect of coulometry is important because the origin, amount, and type of surface carbon present on samples during cleanliness verification analysis may be unknown.

Combustion/Oxidation

Figure 1 shows a schematic representation of the DOCC combustion system. Oxygen gas flows through precombustion scrubbers for the removal of interfering impurities prior to entering a quartz combustion tube. Organic surface carbon is oxidized to CO₂ on samples placed into the 420°C (788°F) heated zone of the quartz combustion tube, generally in about five minutes.

If desired, the samples can be advanced into the 590°C (1094°F) heated zone of the quartz combustion tube, for approximately an additional five minutes, where the inorganic surface carbon is converted to CO₂. The verification and rationale for using these combustion times and temperatures have been recently published.⁷

The 590°C (1094°F) heated zone also contains a combustion catalyst, which converts any incompletely oxidized products (like CO or VOCs) to CO₂ before leaving the quartz combustion tube. After this, the combustion gases are passed through post-combustion scrubbers where interfering gases such as nitrogen and/or sulfur oxides are removed.

Excess oxygen gas sweeps the CO₂ generated from the surface carbon oxidation reactions into the electrochemical cell of the automatic CO₂ coulometer.

Electrochemical Technology

Coulometry is an analytical technique that uses electricity as a means to electrochemically generate and measure the amount of titrant used in a titration.^8,9,10 Apparatus consists primarily of an electrochemical cell, wherein the titration takes place, and the associated electronics used to generate titration current, measure the total charge (coulombs) passing through the cell, and digitally display the results.

Figure 2 shows a schematic representation of the titration cell and CO₂ coulometer. The titration cell consists of a cathode compartment containing a platinum electrode, an anode compartment containing a silver electrode, and the titration solutions. The cathode chemistry is a semi-aqueous solution containing an acid-based indicator and 2-aminoethanol.

As the CO₂ produced from the combustion reaction of surface carbon enters the cathode solution, it reacts quantitatively with 2-aminoethanol to form 2-hydroxyethylcarbamic acid:

CO₂ + NH₂(CH₂)2OH => HO(CH₂)2NHCOOH

2-aminoethanol 2-hydroxyethylcarbamic acid

As the acid forms within the cell solution, the indicator’s color fades from blue to clear. A photometric detector sensing the change in light transmission through the electrochemical cell automatically switches the cell current to "ON." While current flows through the cell, water in the cathode solution is electrolytically reduced at the platinum cathode to form the hydroxide ion titrant:

2H₂O + 2e- => H₂ + 2OH-

As the OH- titrant reacts with the 2-hydroxyethylcarbamic acid analyte, an acid-base neutralization reaction occurs:

OH- + HO(CH₂)2 NHCOOH => H₂O + HO(CH₂)2NHCOO-

titrant analyte

... which restores the cathode solution to its original end-point color, causing the photometric detector to switch the cell current to "OFF."

Faraday’s Law

The amount of hydroxide ion generated and consequently the amount of CO₂ titrated as 2-hydroxyethylcarbamic acid relates to the number of coulombs which pass through the cell during the titration. The weight of carbon titrated is determined by Faraday’s Law:

where eq. wgt. C = 12.011 gms/eq

F = Faraday’s Constant = 96,487 coulombs/eq.

i = current

t = time

Tf = final time at end-point

The instrument electronically measures the total charge and, using the above equation, calculates and digitally displays the weight of carbon detected to the nearest tenth of a microgram. Since the fundamental units of coulombs are measured during this 100-percent efficient coulometric process,⁹ chemical standard calibration is not required.

Sensitivity of the CO₂ analyzer is better than 1 µg C. Overall accuracy is limited by the reproducibility of successive background measurements. Using standard materials, deviations are typically better than 0.15 percent relative or + 1 to 2 µg of C (whichever is greater). Sample homogeneity normally limits the accuracy of the analysis rather than the instrument’s detection capability.¹¹

Cleaned metal coupons, witness panels, small metal parts, or sections of parts can be analyzed, always to be handled with clean tweezers or white linen gloves. Although it’s recommended that samples be analyzed immediately after cleaning, this may not always be possible.

They can be stored in plain manila-style envelopes in a clean desiccator. Storage in common plastic bags is not recommended, as these can be a source of contamination. Leaving cleaned samples exposed to air for prolonged periods causes them to become contaminated with airborne carbon-containing species.

Commercial-grade oxygen gas can be used for the combustion of surface carbon to CO₂, because precombustion oxygen scrubbers are incorporated into the system. Oxygen gas, adjusted to a flow rate of about 200 ml/min., continuously sweeps through the 2.54 cm (1.0 in.) OD quartz combustion tube.

Prior to sample analysis, a background determination is made. The coulometer’s digital display meter is set to zero and, with no sample in the quartz combustion tube, the micrograms carbon per minute are determined. Generally, a background reading of 0.4 µg C/min. or less is obtainable.

Sample Processing

Samples are inserted into the ambient temperature zone of the quartz combustion tube by opening the breach block attached to the tube. The introduction port is sealed and oxygen gas is allowed to sweep out the system for about one minute prior to analysis.

With the coulometric’s digital display meter set to zero, a precombusted hooked stainless steel sample manipulator rod advances the test piece into the 420°C (788°F) heated zone. When CO₂ evolution ceases, generally within five minutes, the digital display reading, which corresponds to the organic portion of surface carbon analyzed, is recorded.

If the surface areas of the samples analyzed are constant, results can be reported simply as:

µg C (analyzed) = µg C (final counts) - µg C (background counts)

Or, if it’s desired to report surface carbon per unit area analyzed, such as milligrams per square meter, then the following formula can be used:

mg C (analyzed)/m² =

[µg C (final counts) - µg C (background counts)] / 1000/surface area analyzed in units of m²

If the analysis of inorganic surface carbon is desired, the samples are advanced in-situ into the 590°C (1074°F) heated zone of the quartz combustion tube with the sample manipulator rod, and the aforementioned process is repeated.

As indicated earlier, instrument calibration is not required. However, the instrument performance can be checked by using sucrose, for example, as a primary standard. A known weight of sucrose placed into a platinum boat is inserted into the quartz combustion tube and analyzed. Users can establish other acceptance criteria that may be more closely related to the type of contaminants or residues (if known) to be present on their samples.

Methods Compared

The DOCC technique has been used to obtain measurements of surface carbon on a variety of metal surfaces.^12,13 The following examples compare DOCC to an indirect and a direct cleanliness verification method, then describe its use for evaluating cleaning procedures, cleaning agents, and manufacturing processes.

An indication sometimes used to assess surface cleanliness is surface wettability. Observing how water flows off a part or test panel can be an indirect method for cleanliness assessment (the "water break" test). If water breaks up into beads and dewets, the part is considered contaminated. If water "sheets" and flows evenly off the part, the part is considered clean.

Quantitatively, surface wettability can be assessed by using a contact angle goniometer to measure the contact angle between a water drop and the surface of a part or test panel that has been cleaned. Large contact angles are associated with poor wettability and contaminated surfaces, whereas small ones are associated with good wettability and clean surfaces.

Illustrated Cleaning

Figure 3 shows the correlation between contact angle measurements and DOCC analyses for different cleaning methods and types of steel. Samples of cold rolled steel, hot-dipped galvanized steel, and electrogalvanized steel were cleaned with either organic solvents or an aqueous alkaline cleaner solution.

Figure 3A shows that, in all cases, the contact angles for samples cleaned with organic solvents were larger than those for the samples cleaned with the aqueous alkaline cleaner. These results imply that the surfaces prepared with the aqueous alkaline cleaner were cleaner than the samples prepared with the organic solvents.

Figure 3B illustrates that the DOCC results correlate directly with the contact angle measurements: Samples that were cleaned with the aqueous alkaline cleaner and exhibited small contact angles also had low surface carbon values.

SIMS Correlation

DOCC measurements have also been compared to SIMS surface chemistry analysis results, which can provide direct information relating to the thickness of a carbonaceous layer present on a sample’s surface.

During SIMS analysis, samples placed in a high-vacuum chamber are bombarded with an energetic primary ion beam. The interaction of this ion beam with the sample’s surface causes surface elemental and molecular fragments to be sputtered away as ionic species. The analysis of these sputtered secondary ions is accomplished with a mass spectrometer. Monitoring the positive carbon ion signal as a function of time results in a depth profile of carbon present on the sample.

Complementary DOCC measurements and SIMS depth profile analyses were performed on three metal samples displaying different surface carbon levels. These results, presented in Figure 4, show that the surface carbon depth profiles correlate well with the DOCC measurements.

Cleaning Procedure Evaluation

In an effort to replace 1,1,1-trichloroethane (TCE) vapor degreasing as a method for cleaning stainless steel tube fittings, unions (with their nuts and ferrules removed) were subjected to the following cleaning protocols:

a. immersed in heated 60°C (140°F) aqueous alkaline cleaner solution with mild agitation;

b. vapor degreased in TCE;

c. ultrasonically cleaned in isopropyl alcohol (IPA) at room temperature; and

d. spray-impinged with heated 60°C (140°F) aqueous alkaline cleaner solution.

After cleaning, the organic surface carbon on the fittings was determined by DOCC. For the purpose of comparison, as-received uncleaned unions were also analyzed.

Figure 5 ranks, compares, and shows the standard deviation of the surface carbon results for the four different cleaning protocols and for the as-received fittings. Results are presented in µg C since the unions all have the same surface area. The as-received unions, supplied in plastic bags and handled manually, showed the highest surface carbon values.

Consistent Results

Immersions with an aqueous alkaline cleaner substantially decreased the surface carbon. However, when compared to the other three cleaning procedures, immersion cleaning with an aqueous alkaline cleaner yielded the highest surface carbon level with the most variation in results. Vapor degreasing using TCE and ultrasonic cleaning with IPA yielded lower but virtually identical results.

The best cleaning procedure, revealing the lowest surface carbon results with the least variability, was spray impingement with an aqueous alkaline cleaning solution.

These results are consistent with the findings of others,^14,15,16 who also showed that equivalent or superior cleaning can be achieved with lower-cost, non-ozone-depleting cleaning agents. Surface carbon results based on DOCC measurements showed that ultrasonic cleaning with IPA and spray impingement cleaning with an aqueous alkaline cleaning solution were found to be equivalent to or better than vapor degreasing with TCE.

Cleaning Agent and Manufacturing Process Evaluation

Using the aforementioned spray impingement cleaning procedure, the cleaning effectiveness of three aqueous alkaline cleaners was compared by DOCC. This evaluation was performed on three steel test panels having identical surface areas but displaying inherently different levels of organic surface carbon.

Cleaning parameters such as dwell time, bath temperature, concentration, and rinse time were held constant. DOCC results shown in Figure 6 indicate that Cleaner 3 was most effective in reducing surface carbon on all three steel surfaces tested.

Designing the manufacturing process to minimize the formation of surface carbon can also be considered as a method for achieving an acceptable surface carbon level.

Table 1 shows surface carbon results obtained by DOCC analysis on two sets of steel samples produced by different manufacturing processes. Not only did Process B leave more than three-and-one-half times more organic surface carbon with greater variability than Process A, it also displayed more than twice as much inorganic surface carbon.

These results, corroborated by other observations gained over time, demonstrated that the DOCC technique was a viable method for relating steel surface cleanliness to the manufacturing process. The origins of surface carbon in the manufacturing process were determined¹⁷ and kept under control by incorporating DOCC as an in-plant statistical process control tool for monitoring surface cleanliness.¹⁸

As industry phases out cleaning solvents that pose potential harm to public health and the environment, designing, developing, and implementing new cleaning processes has become a necessity.

DOCC can play a major role during this mandatory transition period by not only verifying the efficiency of alternative cleaners, but also by keeping new cleaning systems on track as they become part of the manufacturing process.

References

1. Strohmeier, B.R., "Improving the Wettability of Aluminum Foil with Oxygen Plasma Treatments," Contact Angle, Wettability and Adhesion, K.T. Mittal, ed., The Netherlands, VSP, (1993), pp. 453-468.

2. Coduti, P.L., "Effect of Residual Carbon on the Paintability of Steel Strip," Metal Finishing 78, (1980), pp. 51-57.

3. Iezzi, R.A. and Leidheiser, H. Jr., "Surface Characteristics of Cold-Rolled Steel as They Affect Paint Performance," Corrosion 37, (1981), pp. 28-38.

4. Shields, J., Adhesives Handbook, 3rd ed., Butterworths, London, England, (1985), pp. 87-113.

5. Sheppard, K., "Metal Surface Preparation and Cleaning," Electroplating Engineering Handbook, 4th ed., L.J. Durney, ed., Van Nostrand Reinhold Company, New York, (1984), pp. 58-173.

6. Becker, J.F. and Shoemaker, M.C., "An Aqueous Fine Clean Process for Rocket Engineer Hardware," Alternatives to Chlorofluorocarbon Fluids in the Cleaning of Oxygen and Aerospace Systems and Components, ASTM STP 1181, C.T. Bryan and K. Gebert-Thompson, eds., American Society for Testing and Materials, Philadelphia, (1993), pp. 78-92.

7. deVries, J.E., Haack, L.P. and Coduti, P.L., "Measurement of Carbon on Cold Rolled Steel: A Comparative Study Using Surface Analytical and Coulometric Methodologies," Industrial and Engineering Chemistry Research, (in press, 1994).

8. Fisher, R.B. and Peters, D.G., Quantitative Chemical Analysis, 3rd ed., W.B. Sanders Co., Philadelphia, (1968), pp. 752-825.

9. Huffman, E.W.D. Jr., "Performance of a New Automatic Carbon Dioxide Coulometer," Microchemical Journal 22, (1977), pp. 567-573.

10. Johnson, K.M., King, A.E. and Sieburth, J.M., "Coulometric TCO2 Analysis for Marine Studies: An Introduction," Marine Chemistry 16, (1985), pp. 61-82.

11. "Determination of Surface Carbon," Application Note #7, UIC Inc., Joliet, IL, (1994).

12. King, A.E., "Direct Determination of Carbon on Metal Surfaces," The Association for Finishing Processes Technical Paper FC 78-584, Society of Manufacturing Engineers, Dearborn, MI, (1978), pp. 1-9.

13. Coduti, P.L., "Relationship of the Surface Cleanliness and Surface Chemistry to the Corrosion Performance of Painted HSLA Steels for Exposed Autobody Applications," Automotive Corrosion of Deicing Salts, Baboian, R., ed., National Association of Corrosion Engineers, Houston, TX, (1981), pp. 363-376.

14. Marts, K. and Howard, J., "Alternative Cleaning Takes Flight in Aerospace Industry," Precision Cleaning, Nov./Dec., (1993), pp. 35-41.

15. Koch, U.H. and Kmetko, C.J., "Aqueous Cleaning of Industrial and High Purity Valves and Fittings: Process Development, Control and Validation," Alternatives to Chlorofluorocarbon Fluids in the Cleaning of Oxygen and Aerospace Systems and Components, ASTM STP 1181, C.J. Bryan and K. Gebert-Thompson, eds., American Society for Testing and Materials, Philadelphia, (1993), pp. 49-65.

16. Meyers, B., "Cleaning Aluminum Heat Pipe Casings with Replacements for Ozone-Depleting Chemicals," Alternatives to Chlorofluorocarbon Fluids in the Cleaning of Oxygen and Aerospace Systems and Components, ASTM STP 1181, C.J. Bryan and K. Gebert-Thompson, eds., American Society for Testing and Materials, Philadelphia, (1993), pp. 103-115.

17. Coduti, P.L., "Effect of Steel Processing on the Surface Carbon of Cold-Rolled Steel," Technological Impact of Surfaces, American Society for Metals, Metals Park, OH, (1982), pp. 57-101.

18. Pumnea, R.W. and Stadnik, J.M. Jr., "Characterizing Steel Surface Cleanliness Utilizing Statistical Methods for Data Analysis," Second International Symposium on Statistical Process Control and Sensors in the Steel Industry, The Metallurgical Society of the Canadian Institute of Mining and Metallurgy, Montreal, Canada, (1988), pp. 92-124.