Assaying Pentachlorophenol-treated Wood Using XRF

Transcription

1 IRG/WP THE INTERNATIONAL RESEARCH GROUP ON WOOD PROTECTION Section 2 Test methodology and assessment Assaying Pentachlorophenol-treated Wood Using XRF H. M. Barnes 1, C. R. McIntyre 2, D. W. Bullock 3, M. H. Freeman 4, and G. B. Lindsey 1 1 Mississippi Forest Products Laboratory, Forest & Wildlife Research Center, Mississippi State University, Box 9820, Mississippi State, MS 39762, USA 2 McIntyre Associates, 8565 E. Grandview Lake Dr., Walls, MS 38680, USA 3 KMG-Bernuth, 3125 Michelle Dr., Matthews, NC 28104, USA 4 Independent Wood Scientist, 7421 Hunters Tree Cove, Memphis, TN 38125, USA Paper prepared for the 38 th Annual Meeting Jackson Lake Lodge, Wyoming, USA May 2007 IRG SECRETARIAT Box 5609 SE Stockholm Sweden

2 Assaying Pentachlorophenol-treated Wood using XRF H. M. Barnes 1, C. R. McIntyre 2, D. W. Bullock 3, M. H. Freeman 4, and G. B. Lindsey 1 1 Mississippi Forest Products Laboratory, Forest & Wildlife Research Center, Mississippi State University, Box 9820, Mississippi State, MS 39762, USA 2 McIntyre Associates, 8565 E. Grandview Lake Dr., Walls, MS 38680, USA 3 KMG-Bernuth, 3125 Michelle Dr., Matthews, NC 28104, USA 4 Independent Wood Scientist, 7421 Hunters Tree Cove, Memphis, TN 38125, USA ABSTRACT A study was conducted to evaluate the effect of oil type on the performance of test posts treated with oil-penta and exposed in DeQueen, AR for forty years in ground contact. The original assays for the poles were done by lime ignition, a practice used sparingly today in treating plants or by independent inspectors, in part due to the complicated analysis procedure and chances for error. XRF instruments for chlorine analysis have become commonplace. The question put forward by many inspectors, agencies, and treaters is Will XRF accurately predict lime ignition results?. This study encompassed seven different oil systems. Regression analysis yielded an R 2 value over 97% when using XRF to predict lime ignition values. The only caveat was that a proper calibration standard was required for the XRF analysis. Users should be confident that properly derived XRF values will match lime ignition values within 5%. Keywords: pentachlorophenol, hydrocarbon solvent, xray fluorescence spectroscopy, aromatic oil, paraffinic oil INTRODUCTION In the past two decades, the use of x-ray fluorescence spectroscopy (XRF) for the determination of chlorine for analysis of pentachlorophenol has become commonplace. However some treating plants and inspection agencies still use the older Volhard chloride (aka lime ignition) method (AWPA 2006). As part of a larger study, the assay of pole sized stubs was conducted using XRF analysis. In order to compare with the original assays done by lime ignition, a correlation study was conducted. The original southern pine pole stubs were installed in DeQueen, Arkansas in 1960 (Walters 1965) and decommissioned in 2000 (McIntyre et al. 2005). The original study included both air-dried and steam-conditions stubs treated with pentachlorophenol (penta) in seven different oil systems. A brief description of these oils can be found in Table 1. Results from the periodic inspection of these stubs can be found in the literature (Arsenault 1976, McIntyre et al. 2005, Walters 1965, Walters & Arsenault 1971). 2

3 EXPERIMENTAL METHODS A total of 140 southern pine (Pinus echinata) pole stubs, 70 air-dried and 70 steamconditioned were treated with pentachlorophenol in one of the seven different heavy oil hydrocarbon solvents shown in Table 1. Stubs were treated using an initial air pressure of kpa (50-75 psig) followed by a pressure period of 1389 kpa (200 psig) for 1-3 hours. Solution temperatures ranged from C ( F). A final vacuum at kpa (22 in Hg) was applied for hours. Details of the treating of these stubs can be found in the literature (Ferrucci 1962). These stubs were set at a depth of 762 mm (30-inches) in a test plot located in DeQueen, Arkansas in Table 1. Composition and characteristics of treating solutions used. Oil A B C D E F G Composition & Properties (% by vol) Aromatic light fraction Aromatic heavy fraction Paraffinic light fraction Paraffinic heavy fraction Waxy petroleum fraction with high pour point Treating solution concentration (% w/w) Specific gravity, oil 60 F Penta solvency, RT (% w/w) Periodically, the pole stubs were pulled, evaluated for decay and termite attack (see Figure 1), and bored for preservative assay. It is interesting to note the much better Index of Condition (100 = no decay) AIR-DRIED 80 A 75 B C 70 D 65 E F 60 G Year Year Figure 1. Depreciation curves for air-dried (left) and steam-conditioned (right) posts treated with penta in various oils. performance of the steamed posts compared to the air-dried material, presumably due to the sterilization effects of pre-treatment steaming. Index of Condition (100 = no decay) STEAM-CONDITIONED A B C D E F G 3

4 In 2000 when the plot was decommissioned, stubs were removed from the test plot (Figure 2a), evaluated for decay and termite attack (Fig. 2b), and above ground and below ground disks were removed (Fig. 2c, d). These disks were segmented into 13mm (0.5-inch) radial (bark to pith) assay zones for analysis. The data reported by McIntyre et al. (2005) were all obtained at Mississippi State University using XRF analysis according to AWPA (2006) standard A9 using a tube-type x-ray source. Subsequently, samples were sent to an independent third party laboratory for analysis by XRF (also tube-type source) and lime ignition (AWPA 2006). A correlation was run between XRF and lime ignition results in order to develop a prediction curve. a. b. c. d. Figure 2. Procedure for sample collection [(a) removing stubs; (b) evaluating stubs for decay and termites; (c) removing disks for assay; (d) cut stub section]. 4

5 RESULTS AND DISCUSSION A linear regression of lime ignition values (LI) vs. XRF values was established for each of the seven oil carriers in the test and for all oils combined. Regression equations are shown in Figures 3 and 4 for the seven oils tested. The best coefficient of determination (R 2 ) was for Oil B with a composition of 90% aromatic light fraction + 10% aromatic heavy fraction. Oils A, C, and F also gave R 2 greater than 99% (Figure 3). The oil with the worst fit with an R 2 of 91% was the 80/20 mix of paraffinic light fraction to aromatic light fraction (Oil D-Figure 4). Oils E and G yielded coefficients of determination of about 97%. LIME IGNITION (kg/m 3 penta) OIL A: LI = XRF ; R2 = 99.5% OIL B: LI = XRF ; R2 = 99.7% OIL C: LI = XRF ; R2 = 99.1% OIL F: LI = XRF ; R2 = 99.4% XRF (kg/m 3 penta) Figure 3. Regressions for oils A, B, C, and F. 5

6 LIME IGNITION (kg/m 3 penta) OIL D: LI = XRF ; R2 = 91.0% OIL E: LI = XRF ; R2 = 97.4% OIL G: LI = XRF ; R2 = 96.9% XRF (kg/m 3 penta) Figure 4. Regressions for oils D, E, and G. When the data for all oils were combined, an R 2 of almost 98% was obtained as shown in Figure 5. This shows an extremely good fit to the data over a range of oils with widely varying properties. An interesting side issue is shown in Figure 6 which shows the comparison between the two methods of conditioning. It is clear that the method of conditioning did not affect the method of analysis chosen. CONCLUSIONS As part of a long standing study on the effect of oil composition on the durability of penta-treated wood, pole stubs from a forty-year-old exposure site on the border of AWPA Hazard Zones were analyzed for preservative content using both the lime ignition and XRF methods. The resultant correlation curves indicated that XRF analysis 1 See AWPA Deterioration Zone map, Standard U1-06, Figure 1, 2006 Book of Standards, American Wood-Preservers Association, Birmingham, AL, p

7 LIME IGNITION (kg/m 3 penta) OIL A OIL B OIL C OIL D OIL E OIL F OIL G ALL: LI = XRG ; R2 = 97.8% XRF (kg/m 3 penta) Figure 5. Regression for all oils combined LIME IGNITION (kg/m 3 penta) AIR DRIED STEAM-CONDITIONED AD: LI = 1.05 XRF ; R 2 = 97.2% SC: LI = XRF ; R 2 = 98.6% XRF (kg/m 3 penta) Figure 6. Comparison of air-dried and steam-conditioned assay values. 7

8 could be used with great accuracy to predict the original lime ignition values in the study. These results should assure treaters and others of the validity of XRF analysis for quality control and other purposes. Such results are applicable across a wide range of oil types and are valid so long as a proper calibration of the tube-type instrument is done prior to analysis. Interestingly, this study showed that the type of conditioning used should not affect the analysis results. Data from this study are being analyzed and the impact of oil type on the preservative gradient and performance will be presented in a future paper. ACKNOWLEDGEMENTS This paper has been approved as Journal Article FP-408 of the Forest & Wildlife Research Center, Mississippi State University. Grateful appreciation is extended to the AWPA P9 Type A task force of Committee P3-AWPA, the Weyerhaeuser Company, and Timber Products Inspection for their hard work and assistance with the conduct of this project since its inception in REFERENCES American Wood-Preservers Association (2006): Book of Standards. Standard A5-05 Standard Methods for Oil-borne Preservatives; Standard A9-01 Standard Method for Analysis of Treated Wood and Treating Solutions by X-ray Spectroscopy. AWPA, Birmingham, AL. Arsenault, R D (1976): Concentration and distribution of pentachlorophenol in pressuretreated, pine pole stubs and their condition after 15 years exposure, Appendix B to Report of Committee P-3, Organic Solvent Systems. Proceedings, American Wood- Preservers Association 72: Ferrucci, A P Jr (1962): Report of Committee P-3, Petroleum Solutions, Appendix A: Investigaton of properties of petroleum solvents governing behavior of pentachlorophenol-petroleum solutions in wood. Proceedings, American Wood- Preservers Association 58: McIntyre, C R; Barnes, H M; Freeman, M H (2005): 40-Year assays of penta-treated posts in DeQueen, Arkansas. Proceedings, American Wood-Preservers Association 101: Walters, C S (1965): The concentration and distribution of pentachlorophenol in pressure-treated, pine pole stubs following exposure: an ad interim report. Proceedings, American Wood-Preservers Association 61:

9 Walters, C S; Arsenault, R D (1971): The concentration and distribution of pentachlorophenol in pressure-treated, pine pole stubs following exposure. Proceedings, American Wood-Preservers Association 67: