Pressure is the part of this rebuild doing the enabling work. Oxidation in LC-IRMS happens in the flowing liquid, so the team put a back-pressure regulator on the line and held 5.2 MPa across the whole flow path while a modified heater raised the temperature, with sodium persulfate delivered after the column as the oxidant [4]. That is about 52 bar, near 51 atmospheres [15]. Held there, an aqueous stream can be driven roughly 401 degrees past the near-99 C condition where the conventional version of this measurement oxidises [16]. They settled on 500 C as the working temperature [6], and the work appears in Analytica Chimica Acta [2].
The flatness of the recovery window is worth more than the agreement figure. When recovery stays close to 100% anywhere between 300 and 600 C [5], oven temperature is not the parameter that will quietly drift on you mid-run, which is the kind of robustness that separates a method from a demonstration.
Fluorine is a different chemistry problem. The authors say so plainly: the strength and thermal stability of the carbon-fluorine bond left fluorinated compounds with lower recoveries and poorer precision [7]. That constrains the framing, because trifluoroacetic acid is one of the persistent PFAS the paper itself raises as motivation [13]. Chlorinated disinfection by-products are the tractable target today, and they are worth attributing on their own terms, since haloacetic acids are the major DBP class reported to raise bladder cancer risk [12].
Then there is the denominator. Measurement became possible only at 500 mg/L and above under the tested conditions [8], which is half a gram of analyte per litre of sample [14]. The paper does not compare that floor with ambient environmental levels, so the thing it does not tell you is how many stages of preconcentration sit between this interface and a real water sample. On the evidence here, sample preparation is the bottleneck that decides whether this method ever sees a field sample.
Why chase isotopes at all: routine monitoring answers how much and stops there, offering little about sources or transformation [10], while carbon isotope ratios carry origin and production history, and the same compound made by two routes can carry two different values [9]. That is the same logic already used to catch food fraud [9]. Kawashima describes the system as a new tool for source identification and fate analysis [17], which is an honest description of an instrument paper.
My reading: believe it for chlorinated and brominated acids in concentrated solution, and hold the PFAS attribution claim until the fluorine recoveries move.
