Experimental assessment of the purity of α-cellulose produced by variations of the Brendel method: Implications for stable isotope (δ13C, δ18O) dendroclimatology

Stable isotope dendroclimatology using alpha-cellulose has unique potential to deliver multimillennial-scale, sub-annually resolved, terrestrial climate records. However, lengthy processing and analytical methods often preclude such reconstructions. Variants of the Brendel extraction method have reduced these limitations, providing fast, easy methods of isolating alpha-cellulose in some species. Here, we investigate application of Standard Brendel (SBrendel) variants to resinous soft-woods by treating samples of kauri (Agathis australis), ponderosa pine (Pinus ponderosa) and huon pine (Lagarastrobus franklinii), varying reaction vessel, temperature, boiling time and reagent volume. Numerous samples were visibly 'under-processed' and Fourier Transform infrared spectroscopic (FTIR) investigation showed absorption peaks at 1520 cm(-1) and similar to 1600 cm(-1) in those fibers suggesting residual lignin and retained resin respectively. Replicate analyses of all samples processed at high temperature yielded consistent delta C-13 and delta O-18 despite color and spectral variations. Spectra and isotopic data revealed that alpha-cellulose delta C-13 can be altered during processing, most likely due to chemical contamination from insufficient acetone removal, but is not systematically affected by methodological variation. Reagent amount, temperature and extraction time all influence delta O-18, however, and our results demonstrate that different species may require different processing methods. FTIR prior to isotopic analysis is a fast and cost effective way to determine alpha-cellulose extract purity. Furthermore, a systematic isotopic test such as we present here can also determine sensitivity of isotopic values to methodological variables. Without these tests, isotopic variability introduced by the method could obscure or 'create' climatic signals within a data set.