Capping the hydroxyl groups (-OH) of α-cellulose to reduce Hy-groscopicity for accurate 18O/16O measurement by EA/Py/IRMS

Obtaining an accurate measurement of O-18/O-16 at natural abundance level for land plants-derived a-cellulose with the currently popular EA/Py/IRMS (elemental analysis/pyrolysis/isotope ratio mass spectrometry) method is a challenge due to the hygroscopic nature of the exposed hydroxyl groups, as the O-18/O-16 of adsorbed moisture is usually different from that of the a-cellulose and the relative amount of adsorbed moisture is sample-and relative humidity-dependent. To minimize the hygroscopicity-related measurement error, we capped the hy-droxyl groups of a-cellulose by benzylation to various degrees and found that the O-18/O-16 ratio of a-cellulose increased with the degree of benzyl substitution (DS), consistent with the theoretical prediction that a reduced presence of exposed hydroxyl groups should lead to a more accurate (and therefore more reliable) a-cellulose O-18/O-16 measurement. We propose the establishment of a moisture adsorption-degree of substitution or per-centage of oxygen-O-18/O-16 ratio equation, based on the measurement of C%, O% and d(18)O of variably capped a-cellulose, so that a robust correction can be made in a plant species-and laboratory conditions-specific manner. Failure to do so will lead to an average underestimate of a-cellulose d(18)O by 3.5 mUr under "average" laboratory conditions.