SNO-STR-97-001 January 9, 1997. Tests of acrylic with phenol solutions. B. Knox and C. Goodwin. 1. Introduction and first series of tests. Aromatic hydrocarbons, that is those containing one or more benzene rings, are reported to attack acrylic (PMMA or polymethylmethacrylate) if undiluted or in concentrated solutions. But there seems to be very little information available on the effects of very low concentrations in aqueous solutions. Such very dilute solutions could arise in several possible ways. One way is from the decomposition products of tetraphenylboron (TPB) used to remove K from MgCl2 solutions. Diphenylboric acid (DPBA) and biphenyl (also known as diphenyl) are reported to be the most common decomposition products. But triphenylboron (TriPB) and phenol can be formed by another decomposition route and both DPBA and TriPB are subject to further decomposition. (Douglas B. Hunter and Paul M. Bertsch, Environ. Sci. Technol., (1994) vol.28, 686-691). Other possible ways considered were the leaching of phenolic compounds from phenolic-cured butyl rubber O-rings and even the attrition and degradation of poly(styrene-divinylbenzene) ion-exchange beads. Since these dilute solutions could possibly come into contact with the Acrylic Vessel of the SNO detector it is necessary to determine if any harmful effects are possible. Phenol was selected as a representative of such compounds. Solutions were prepared in 100ml Duran glass reagent bottles with strengths of 1%, 0.1%, 0.01%, 0.001% (10 ppm) and 0.0001% (1 ppm). Standard Laboratory Reagent grade phenol (Fisher Scientific) and UPW were used to prepare the solutions. Into each bottle were placed 3 cuvettes made from high-purity UV-transmitting PMMA (Kartell, Italy), using plastic tweezers to avoid hand contact. These cuvettes are of course made by injection moulding, using a moulding grade of PMMA. The chemical resistance is therefore expected to be slightly inferior to that of a high-molecular-weight PMMA made by casting, such as that being used for the AV. In addition the cuvettes are hollow and thus have 4 air-acrylic interfaces in the path of a transmitted beam, thus making surface attack easier to detect. The dimensions of these cuvettes are 10 mm x 10 mm internal and 12 mm x 12 mm external, thus giving a total path length of only 2 mm of acrylic in the beam. The time was the late evening of December 2nd, 1996. The five bottles were stored away from direct light at about 18C. After 3.5 days one cuvette was removed from each bottle, rinsed in RO water and allowed to dry. There were no obvious effects detected by visual inspection. Unfortunately our Shimadzu UV-1201 UV/Vis spectrophotometer is not working properly at the moment and is awaiting repair. The cuvette which had been immersed in 1% phenol did seem to have a faint smell of phenol coming from it, suggesting that there might have been some absorption. These cuvettes were later examined (by CG) using a Pye Unicam PU 8800 spectrophotometer in the Thin Films Facility in this department, and compared with a pristine cuvette which had not been immersed in any solution. The transmission was measured from 850 nm to 190 nm for two cuvettes, the one which had been immersed in 1.0% phenol and the pristine cuvette. All six cuvettes were then scanned over the same range using an expanded (80% to 90%) transmission sensitivity. The recorder trace differences between the cuvettes were small and not conclusive, but the transmission of the pristine cuvette seemed to be slightly higher than those which had been immersed. At 420 nm, where the difference appeared to be greatest, the transmission of the pristine cuvette was about 84.5% while that of the cuvettes which had been immersed in phenol solutions was about 84%. However the reproducibility of transmission of these disposable cuvettes is only guaranteed to 1% and the pristine cuvette was dry; a comparison with one which had been immersed in UPW only would be a better test. There was also an effect on the short-wavelength cut-off point. The 10% transmission point seemed to be shifted by about 22 nm. See figure 1. Again a better comparison would have been with a cuvette which had been immersed in UPW only. However the one immersed in 1 ppm phenol solution may be a quite reasonable approximation to this. 2. Second series of tests, using stronger solutions. Since the effects of dilute solutions were so small over a short time like 3.5 days a stronger solution (5% phenol) was prepared. At 18 C the solubility of phenol in UPW is about 6%, so this was a quite concentrated solution. Two cuvettes were placed in this solution on 20th December 1996. After one day there was no visible effect. After 2 days there was a definite haziness and loss of clarity. After 6 days both cuvettes had softened and slumped into opaque white rubbery blobs. They were removed from the solution and stored in polythene bags. Transparency was slowly restored, presumably as water vapour was lost, but the blobs remained soft and rubbery. Since the effect of a 5% solution was so drastic an intermediate strength solution (3% phenol) was prepared. Two cuvettes were immersed in this solution on 26th December 1996. After 1 day and again after 2 days there were no detectable effects by visual inspection. After 3 days a slight loss of clarity was detected, and after 4 days this loss of clarity was readily detectable although still small. After 5 days the cuvettes were distinctly hazy in appearance, and after 6 days this haziness was pronounced. After 7 days the cuvettes were too hazy to easily see through and were taken out. They were slightly tacky and tended to stick together. They were rinsed with UPW and allowed to dry. As they dried the haziness decreased and they became transparent again. Both these cuvettes were later examined (by CG) using the PU 8800 spectrophotometer, again scanning from 850 to 190 nm, and compared with the pristine cuvette. There was a small reduction in the measured transmission of the cuvettes which had been immersed over the range from 850 nm to about 340 nm. But this was only about 1% and could be partly due to absorbed water. However there was a more significant reduction below about 340 nm and the measured wavelength at which there was 10% transmission was 239 nm for the pristine cuvette and 261 nm for the cuvettes which had been immersed in 3% phenol solution for 7 days, a difference of 22 nm. See figure 2. 3. Effects of longer-term immersion in dilute solutions. After 28 days of immersion a second cuvette was removed from each of the five bottles of dilute phenol solutions, rinsed in UPW and allowed to dry. No obvious effects could be detected by visual inspection of the dry cuvettes, although a very slight loss of clarity of the cuvette in 1% phenol had been suspected before it was removed from the solution. This set of 5 cuvettes was also examined (by CG) using the PU 8800 spectrophotometer. There was no significant difference in the transmission measured for these cuvettes over the range from 850 nm down to 340 nm. But below 340 nm the transmission of the cuvette which had been immersed in 1% phenol solution was significantly below that of the other cuvettes. None of the other cuvettes showed any significant differences from each other. See figure 3. The wavelength at which 10% transmission is measured is displaced by about 20 nm for the cuvette which had been immersed in 1% phenol solution. The practical effects of reduced transmission below 340 nm would be small, since the transmission of the AV is expected to be negligible below about 300 nm. It is the fact that a detectable effect has been observed which is significant. 4. Conclusions and recommendations. PMMA cuvettes have good resistance to dilute solutions of phenol, and solutions of 0.1% and less seem to have no significant effect on the optical clarity or light transmission of PMMA. But a solution of 1% phenol does have a significant effect, and they are attacked and softened by more concentrated solutions. The effects of phenol solutions on the AV are likely to be similar to those reported here, although probably not identical since the grade of PMMA is not the same. It may be desirable to carry out tests on samples of the actual material being used to construct the AV, but this work indicates the concentrations at which significant effects may be expected. This work has only looked for optical effects. The possible effects on the mechanical properties have not been investigated. One point which should be considered is that phenol is not very volatile (b.p. 182 C) and thus a dilute solution which is allowed to dry onto an acrylic surface is expected to become a concentrated solution as it dries. Apart perhaps from benzene, which is more volatile than water, this behaviour is likely to occur for all aromatic compounds and also for detergents, some of which may contain aromatic compounds. Manufacturers of plastics normally recommend that items which have been washed with detergents should be rinsed well before being dried. In addition, many aromatic compounds are much less water-soluble than phenol and so would be deposited from less concentrated solutions. Precipitation of relatively insoluble aromatic compounds from an aqueous solution could perhaps lead to localised attack at the waterline. The AV chimney has two waterlines, one internal (D2O) and one external (H2O). Localised attack at either would be equally undesirable. Fortunately the large volumes involved (kiloton or more) mean that significant amounts of even the most insoluble compounds must be present before the solubility limit could be reached, but it is recommended that all possible precautions should be taken to ensure that in any aqueous solutions which contact the AV the concentrations of aromatic compounds, detergents, surfactants or organic solvents should not only be low but should also be well below the solubility limit in each case. Barrie Knox, SNO group. Chris Goodwin, Thin Film Facility. Copies of the figures will be faxed to H. Lee and M. Shatkay, and to anyone else who requests them.