Long-term immersion tests of "Omega" coated aluminium sheet as used for concentrators. SNO-STR-98-012 B. Knox. 2nd July 1998. 1. Introduction. It was realised in 1991 that all systems using large amounts of pure or ultra-pure water were still liable to experience problems caused by the growth of bacteria, algae and other micro-organisms and that SNO was not likely to be an exception to this. Information was available in the literature at that time, mainly based on the experience gained by the semiconductor and pharmaceutical industries, and there was some information on which biocides had been found most useful for treating such problems. See for example the article by M.W. Mittelman, pp.30-40 and 70, in the issue of "Microcontamination" for January 1986. It was decided to put samples of acrylic and of the reflectively- coated aluminium sheet (then being obtained under the trade name of Omega and being tested for use in the light-concentrating cones to be fitted to the PMT's in SNO) into sample solutions of various biocides that might be considered for use. The acrylic used for these tests was "Oroglas", made by Rohm and Haas, since the acrylic to be used for the AV had not been selected at that time. Similarly the coated aluminium sheet which was actually used for the petals of the concentrators was obtained from the company trading as OCLI. Nevertheless it was felt that the properties of the materials being tested would be similar enough to those finally chosen that the results would be useful. 2. Types of biocide and their use. Biocides can be grouped into a number of types which have some properties in common. a). There are the oxidising types such as chlorine, ozone, chlorine dioxide, hydrogen peroxide and iodine. b). There are quaternary ammonium compounds and related compounds. c). There are aldehydes such as formaldehyde and glutaraldehyde. d). There are anionic and nonionic surface-active agents. There are other materials which can be used as biocides or as additives to enhance the performance of a mixture, but the above are the main types which seemed at that time to be most used in purified-water systems. The low nutrient levels found in UPW systems have the effect of causing free-floating bacteria to be few. Most of the micro-organisms are found attached to surfaces as biofilms. More is now known about the great complexity which these biofilms can exhibit. See for example the article "Slime City" by Andy Coghlan, in New Scientist, 31 August 1996. Effective treatment may require not only a suitable biocide used at a concentration sufficient to give a complete kill but also some kind of surface-active agent to enhance penetration of the biofilm and to help to dislodge it from the surface. And as much as possible of the chemical residues and dead micro-organisms must be flushed away in order to prevent them being used as a source of nutrients in a regrowth of the biofilm or acting as a source of contamination of the purified water. In the case of SNO, the possibility of such residues absorbing traces of radioactivity and later releasing radon or other radioactive isotopes is also a matter for concern. 3. Experimental. The following chemicals were used for these tests. Sodium hypochlorite solution, GPR, from BDH. About 14% available chlorine. Recommended dilution around 0.1%. Adesol 20. This is dioctyl dimethyl ammonium chloride. At the time made by BDH but now made by Coalite. Suggested dilution 1:4000 for pure water systems. Benzalkonium chloride. This is alkylbenzyldimethyl-ammonium chloride. It was supplied by Aldrich. Chloramine-T. This is N-chloro-p-toluene-sulfonamide, sodium salt. Also supplied by Aldrich. Suggested dilution about 500 ppm. Chloral hydrate, GPR, from BDH. Suggested dilution about 0.1%. Formalin solution, phosphate-buffered, pH 7.2 - 7.4. BDH-Gurr. Ready for use. Actual strength not known and probably could have been diluted. Glutaraldehyde, 50% solution, GPR, BDH. Used diluted 1:10 to give a 5% solution. Hydrogen peroxide solution, GPR, BDH. About 30% w/v H2O2, or "100 volumes". Used diluted 1:10. Gerrard A.S.A.B. Composition not known. Supplied by Fisons, since taken over by Fisher Scientific. Not in current Fisher catalogue. Recommended dilution 0.2% to 1%. "Kleeneze" quaternary disinfectant. Composition not known. Was supplied by Fisons but is in current Merck catalogue. Recommended dilution 1:100. The test-pieces were either "petals" taken from prototype trial production (but not bent) or pieces of similar size cut from sheet samples. They were placed in numbered glass bottles filled with the various biocide solutions. Ambient temperature was around 20-22 C. 4. Results. These test were started during the autumn of 1991, and the tests on the Oroglas samples have already been reported. See SNO-STR-97-024. In addition related tests on acrylic (PMMA) in the form of spectrophotometer cuvettes and using phenol solutions of various strengths have also been reported. See SNO-STR-97-001. In general it was found that biocides used at their recommended dilutions did not appear to have any significant effects on acrylic even after prolonged immersion. A preliminary examination of the test-pieces was carried out on 8th December 1991. At that time none of the Oroglas samples showed any detectable effects and only the Omega samples were closely examined. The results are detailed below, as Table 1. Table 1. ________ Bottle Start No. of Biocide solution. Condition of number. date. days. test-piece. 16. 28-8-91 95 Adesol 20, 1:100 Slightly dulled, bluish tinge. 21. 5-9-91 87 Formalin, phosphate- Good. buffered, used as recd. 22. 5-9-91 87 Adesol 20, 1:400 Good. 23. 5-9-91 87 Adesol 20, 1:4000 Good, maybe very good. 24. 5-9-91 87 Benzalkonium Slightly dulled. chloride, 0.22% 25. 5-9-91 87 Chloramine-T Very slightly dulled, and hydrate, 0.18% a white spot formed at one point on cut edge. 32. 14-9-91 78 Chloral hydrate, 0.4% Very good. 33. 14-9-91 78 Glutaraldehyde, 5% Very good. 35. 14-9-91 78 Hydrogen peroxide, 10% Good, but bubbles formed on cut edges. 37. 18-9-91 74 Sodium hypochlorite, Surface very badly pitted 0.7% (0.1% Cl) and etched. Back white. 39. 18-9-91 74 Sodium hypochlorite, Surface lightly pitted 0.07% (0.01% Cl) and etched. Back white. 42. 4-11-91 41 Gerrard A.S.A.B, 0.2% Very good. 43. 4-11-91 41 Gerrard A.S.A.B, 1% Very good. 44. 4-11-91 41 Kleeneze quaternary Very good. disinfectant, 1% From these preliminary results it was quite clear that sodium hypochlorite, which is widely used as a bleach, was totally unsuitable for use on or near the concentrators. Even before this examination relatively rapid attack of the reflective surface and cut edges had been noted, and experiments in Canada had given very similar results. It should be pointed out that these test-pieces did not have any edge protection or evaporated coating such as was applied to the production petals. Some of the other biocides had shown signs of causing dulling or slight attack, and it was noticeable that these all contained chlorine. The samples were then stored at room temperature (20-22 C) for a period of over six years. When the cavity light water fill was started it was decided that they should be taken out and examined again. The petals were removed from their bottles, rinsed with RO water and then allowed to drain and dry. The date was 16th June 1998. Results are given below as Table 2. Table 2. ________ Bottle Start Biocide solution. Condition of number. date. test-piece. 16. 28-8-91 Adesol 20, 1:100 Sample badly attacked, in a few places eaten away, yet most of surface still more or less reflective. 21. 5-9-91 Formalin, phosphate- Sample only slightly buffered, used as recd. dulled. 22. 5-9-91 Adesol 20, 1:400 Sample very badly attacked, eaten away in one place. 23. 5-9-91 Adesol 20, 1:4000 Sample very badly attacked. Eaten away in one spot. 24. 5-9-91 Benzalkonium Sample very badly attacked. chloride, 0.22% 25. 5-9-91 Chloramine-T Sample very badly attacked. hydrate, 0.18% Eaten away in a couple of places. 33. 14-9-91 Glutaraldehyde, 5% Sample badly attacked. 37. 18-9-91 Sodium hypochlorite, Sample very badly attacked. 0.7% (0.1% Cl) 39. 18-9-91 Sodium hypochlorite, Sample very badly attacked. 0.07% (0.01% Cl) 42. 4-11-91 Gerrard A.S.A.B, Sample very good, with no 0.2% significant attack. 43. 4-11-91 Gerrard A.S.A.B, Sample very good, with no 1% apparent attack. 44. 4-11-91 Kleeneze quaternary Sample very good; possibly disinfectant, 1% very slight dulling. 5. Conclusions. The tests carried out at Oxford, together with other experiments carried out at UBC and at CRPP and at site, have shown that care must be taken when choosing a biocide or cleaning agent in order to ensure that it is compatible with all the SNO detector components with which it may come into contact. The coated aluminium petals used in the concentrators have been shown to be one of the most sensitive items present. They are quite rapidly attacked by sodium hypochlorite and biocides based on or containing compounds of this type should not be used. Many organic biocides which contain chlorine might be usable for short contact times but it would be unwise to use them for more extended contact times. Two proprietary products do appear to be usable for long contact times, Gerrard A.S.A.B and "Kleeneze" quaternary disinfectant. But the compositions and manufacturers of both products are not known at this time, and nor is their availability in Canada. It is of course very likely that both products and similar ones can be obtained in Canada. Relevant earlier SNO reports on methods of treating biological activity during or after water-fill of the cavity include SNO-STR-92-010, SNO-STR-92-057 and SNO-STR-93-007, by C. Waltham and others.