Improving memory effect correction to achieve high?precision analysis of ? ¹⁷ O, ? ¹⁸ O, ? ² H, ¹⁷ O?excess and d?excess in water using cavity ring?down laser spectroscopy

Rationale The precision obtained in routine isotope analysis of water (?17O, ?18O, ?2H, 17O-excess and d-excess values) using cavity ring-down spectroscopy is usually below the instrument specifications provided by manufacturer. This study aimed at reducing this discrepancy, with particular attention paid to mitigating memory effect (ME). Methods We used a Picarro L2140i analyzer coupled high-precision A0211 vaporizer an A0325 autosampler. magnitude duration ME were estimated 24 series 50 successive injections samples contrasting compositions. Four correction methods compared, performance was evaluated over 17-month period analysis, two different run architectures. Results remains detectable after 30th injection, implying that common procedures only based on last preceding sample need be revised. developed new composition several samples, designed architecture minimize ME. standard deviation from measurement quality assurance seven-month 0.015‰ for ?17O, 0.023‰ 0.078‰ 0.006‰ 0.173‰ d-excess. In addition, we first ?17O values GRESP certified reference material. Conclusions demonstrates long-term persistence ME, which often overlooked natural samples. As already evidenced when measuring labelled water, it calls consideration compositions previous obtain appropriate correction, prerequisite achieve data.