Picomolar‐scale compound‐specific isotope analyses

• Published in: Rapid communications in mass spectrometry 2018-05, Vol.32 (9), p.730-738
• Main Authors: Baczynski, Allison A., Polissar, Pratigya J., Juchelka, Dieter, Schwieters, Johannes, Hilkert, Andreas, Summons, Roger E., Freeman, Katherine H.
• Format: Article
• Language: English
• Subjects: Accuracy; Alkanes; Astrobiology; Carbon; Carbon dioxide; Carbon isotopes; Chromatography; Dilution; Firmware; Gas chromatography; Low concentrations; Organic chemistry; Resolution; Sensitivity enhancement; Transfer lines
• ISSN: 0951-4198; 1097-0231
• DOI: 10.1002/rcm.8084

Rationale We report modifications to compound‐specific isotope analyses (CSIA) to enable high‐precision isotopic analyses of picomoles of carbon for intact organic molecules. This sample size is two orders of magnitude below the amounts required for commercial systems. The greatly enhanced sensitivity of this system expands molecular isotope studies and applications previously prohibited by low concentrations and small samples. Methods We utilize the resolving power and low volumetric flow rates of narrow‐bore capillary gas chromatography to improve sample transfer efficiency while maintaining narrow peak widths. Post‐column peak broadening is minimized using a micro‐fluidic valve for solvent diversion, capillary combustion reactor, narrow‐bore capillary transfer lines, and cryogenic water trap. The mass spectrometer was fitted with collector amplifiers configured to 25 ms response times and a data logger board with firmware capable of rapid data acquisition. Carbon dioxide gas was introduced directly into the ion source to evaluate the dynamic range of the system and accuracy and precision of carbon isotope ratio (δ13C value) measurements. The accuracy and precision for combusted compounds were evaluated using a suite of n‐alkanes. Results For ≥30 pmol carbon introduced directly into the ion source, the mean difference between the measured and expected δ13C values is 0.03‰ (1σ, n = 57) and the standard deviation of replicate measurements is 0.11‰ (1σ). The CO2 peak widths generated by the exponential dilution flask were 250 ms and the peak widths produced by combusting n‐alkanes were ca 500 ms, less than 25% the width of conventional gas chromatography peaks. For a mixture of 15 n‐alkanes (n‐C16 to n‐C30), the accuracy is 0.3‰ (1σ) and precision is 0.9‰ (1σ) for replicate δ13C measurements with 100 pmol carbon per compound on column. Conclusions The pico‐CSIA method described here offers improved chromatographic resolution and reduces sample size requirements by two orders of magnitude. These advances significantly broaden the available analytical window for CSIA in research areas frequently hindered by sample size limitations, such as forensics, paleoclimate, astrobiology, and biochemistry.