Free amino acids quantification in cloud water at the puy de Dôme station (France)

Abstract. Eighteen free amino acids (FAAs) were quantified in cloud water sampled at the puy de Dôme station (PUY – France) during 13 cloud events. This quantification has been performed without concentration neither derivatization, using LC-MS and the standard addition method to avoid matrix effects. Total concentrations of FAAs (TCAAs) vary from 1.2 µM to 7.7 µM, Ser (Serine) being the most abundant AA (23.7 % in average) but with elevated standard deviation, followed by Glycine (Gly) (20.5 %), Alanine (Ala) (11.9 %), Asparagine (Asn) (8.7 %), and Leucine/Isoleucine (Leu/I) (6.4 %). The distribution of AAs among the cloud events reveals high variability. TCAA constitutes between 0.5 and 4.4 % of the dissolved organic carbon measured in the cloud samples. AAs quantification in cloud water is scarce but the results agree with the few studies that investigated AAs in this aqueous medium. The environmental variability is assessed through a statistical analysis. This work shows that AAs are correlated with the time spent by the air masses in the boundary layer, especially over the sea surface before reaching the PUY. The cloud microphysical properties fluctuation does not explain the AAs variability in our samples confirming previous studies at PUY. We finally assessed the sources and the atmospheric processes that potentially explain the prevailing presence of certain AAs in the cloud samples. The initial relative distribution of AAs in biological matrices (proteins extracted from bacterial cells or mammalian cells, for example) could explain the dominance of Ala, Gly and Leu/I. AA composition of aquatic organisms (i.e., diatoms species) could also explain the high concentrations of Ser in our samples. The analysis of the AAs hydropathy also indicates a higher contribution of AAs (80 % in average) that are hydrophilic or neutral revealing the fact that other AAs (hydrophobic) are less favorably incorporated into cloud droplets. Finally, the atmospheric aging of AAs has been evaluated by calculating atmospheric lifetimes considering their potential transformation in the cloud medium by biotic or abiotic (mainly oxidation) processes. The most concentrated AAs encountered in our samples present the longest atmospheric lifetimes and the less dominant are clearly efficiently transformed in the atmosphere, potentially explaining their low concentrations. However, this cannot fully explain the relative contribution of several AAs in the cloud samples. This reveals the high complexity of the bio-physico-chemical processes occurring in the multiphasic atmospheric environment.


ND: Not determined LOQ: Limit of Quantification (≈ standard deviation, see Figure S3, Table S3 and Section 3.1) Unlike Table S3, negative values are considered as below the LOQ. ND: Not determined LOQ: Limit of Quantification (≈ standard deviation, see Figure S3, Table S3 and Section 3.1) Unlike Table S3, negative values are considered as below the LOQ.  Figure S3, Table S3 and Section 3.1) Unlike Table S3, negative values are considered as below the LOQ.  Figure S3, Table S3 and Section 3.1) Unlike Table S3, negative values are considered as below the LOQ.    Table S3. Concentration (µg L -1 with dilution 9:1, detailed in Figure S3), calibration curve and R² data for the 18 amino acids (AA) analyzed in the 13 clouds sampled at PUY. The calculation method (detailed in Figure S3) might mathematically provide negative values for the concentration. However, if the concentration (Conc) may turn out to be positive due to a higher STD (STD > |Conc|), the values are left as is (e.g., Asn -1 ± 4). Otherwise (STD < |Conc|), we assume to be below the limit of quantification (< LOQ). ND: Not Determined.  Gly > Ala = Pro Zhu et al. (2021) Estimated lifetimes of AAs : Description of the calculations performed in Table 4. Table 4)

1-Calculations of the lifetimes considering theoretical HO • , O3 and 1 O2 * concentrations (column (A) in
Aqueous concentrations of HO • , O3 and 1 O2 * are respectively equal to 10 -14 , 5.0 10 -10 and 1.0 10 -12 M. The concentration of HO • derives from the study of Arakaki et al. (2013); the concentration of O3 is calculated considering a 50 ppb concentration of gaseous O3 and its Henry's law constant (H(O3) = 10 -3 M atm -1 ). 1 O2 * concentration is estimated to be 2 orders of magnitude more concentrated than HO • . All the kinetic constants derive from the Jaber et al. (2021) study (considering T and pHdependency when necessary and available). The lifetimes for individual AA are calculated as following: Table 4)  Table 2 in Jaber et al., 2021). For Arg, Asn, Asp, Gln, Gly, Lys and Pro, lifetimes cannot be calculated since a production is observed during the experiment.

2-Calculations of the lifetimes using irradiation experiments in artificial cloud medium (column (D) in
The lifetimes for individual AA are calculated as follows:

Quantification and uncertainty (Figure S3)
In standard addition, known quantities of analyte (AA) are added to the unknown quantity in the sample. From the increase in signal, we deduce how much analyte was originally in the sample. This method requires a linear response to analyte (Broekaert, 2015).
The magnitude of the intercept on the x-axis is the original concentration of Gly. The equation of the trendline is y = a x + b. The x_intercept is obtained by setting y = 0: x = -b / a, with a = slope of the curve, b = y_intercept, x = the concentration of the AA, y= the mass spectral area: Gly: a = 410.49; b = 13607 → |x_intercept| = [Gly] = 33.1 µg L -1 (negative value) The obtained values are then corrected by the dilution factor of 10 % (due to the ratio 9:1 volume cloud: volume added standard). Final value is: [Gly]= 33.1 × %: ; = 36.8 µg L -1 .
The uncertainty in the x_intercept is sx: where a is the absolute value of the slope of the trendline, n is the number of data points, ? @ is the mean value of y for the points, D = are the individual values of D, D̅ is the mean value of y for the points, and : < is the standard deviation for y: