1 2 3 4 5 6 Supporting information Online HPLC-ESI-HRMS Method for the Analysis and Comparison of Different Dissolved Organic Matter Samples Claudia Patriarca a, *, Jonas Bergquist a, Per J. R. Sjöberg a, Lars J. Tranvik b, Jeffrey A. Hawkes a 7 8 9 a Department of Chemistry BMC, Analytical Chemistry, Uppsala University, Uppsala, Sweden b Department of Ecology and Genetics, Limnology, Uppsala University, Uppsala, Sweden 1 11 * Corresponding author Phone +46 18 471 36 79, e-mail: claudia.patriarca@kemi.uu.se. 12 13 14 15 16 17 18 19 21 22 23 24 25 26 27 28 29 CONTENT 1 Number of pages: 1 2 Supplementary Material and Methods 3 Figures and tables - Figure S1: SRFA and blank chromatograms complete with employed gradients - Table S1: HPLC solvent composition for the samples analysis - Figure S2: Comparison between LC and DI results from the marine sample - Figure S3: Comparison between LC and DI results from the lake sample - Figure S4: Method reproducibility tested at different loading volumes - Figure S5: Comparison of length of CH 2 homologous series - Figure S6: Percentage of N-containing formulas for the three different sources - Figure S7: N-containing formulas distribution between marine and lake samples 4 Supplementary Excel file (Tables.xlsx) - Bray-Curtis dissimilarity (%): dissimilarity matrix of the whole data set expressed as percentage - Pearson s corr. coefficients: covariance coefficients of the whole data set - Contaminants list: list of the contaminants removed from the mass list S1
3 MATERIALS AND METHODS 31 32 33 34 35 36 37 38 39 41 42 43 Reagents and materials All reagents were of analytical or high-performance liquid chromatographic grade: Methanol (LiChrosolv, Merck, Germany), acetonitrile (Sigma-Aldrich, Germany), ultrapure water (VWR chemicals, Germany) and formic acid (HiPerSolvChromanorm 99% VWR chemicals, Germany) high purity hydrochloric acid (3%, Suprapur, Merck, Germany). High-purity water was produced in the lab by a Milli-Q System Millipore (Molsheim, France). Suwannee River Fulvic Acid (SRFA, 2S11F) reference material was purchased from the International Humic Substance Society (Minnesota, United States). Three model compounds with molecular formula C 16 H 18 O 1 and deprotonated mass 369.8272 were purchased, two from Sigma Aldrich (7,8-Dihydroxy-6-methoxycoumarin and 2-(4-(2,2-Dicarboxy-ethyl)-2,5- dimethoxy-benzyl)-malonic acid) and one from Toronto Research Chemicals (Isoferulic Acid 3-O-β-D-Glucuronide). Glassware was heat treated at 45 ºC for at least 4 hours prior to use to remove organic contamination. 44 45 46 47 48 49 5 51 52 53 Sample collection and pre-treatment In this study, three samples from different natural sources were considered. The first was taken 674 m depth from the North Pacific Ocean at the Natural Energy Laboratory of Hawaii Authority (NELHA). 1 This is referred to as the marine sample and is taken from one of the oldest water masses in the ocean. The second source was a brown water lake, named Plåten, located in Sweden (59.8627º N, 18.5426 E), and the last sample was from the reference material Suwannee River Fulvic Acid (SRFA). Marine and lake samples were acidified (.1 M HCl), de-salted and concentrated by solid phase extraction on divinyl-benzene adsorber (PPL, Varian) to a final concentration in methanol of and 35 ppm DOC, respectively 2. A more detailed description of this S2
54 55 procedure is described elsewhere 1. These samples were stored at - C until the day of analysis. 56 57 58 59 61 62 63 64 65 66 67 68 69 7 Sample preparation for the HPLC and DI analysis An aliquot from of the marine sample SPE eluate was diluted with ultra-pure water, enriched with.1% formic acid to a final concentration of 5 ppm C in 5% methanol, 94.9% water,.1% formic acid. The methanol could not be completely removed without precipitation of the part of the sample. The lake sample extract aliquot was evaporated under a gentle nitrogen stream to almost complete dryness and diluted again to 5 ppm C with 5% acetonitrile, 94.9% water,.1% formic acid. The freeze-dried SRFA powder was weighed and diluted to 5 ppm C with 5% acetonitrile, 94.9% water,.1% formic acid. These solutions were transferred into 25 μl insert vials and analyzed by reverse-phase HPLC-MS. The sample preparation for the direct infusion (DI) analyses involved an additional step. Aliquots of the solution analyzed by liquid chromatography were diluted with acetonitrile in order to reach the required solvent composition (acetonitrile- water, 1:1) and DOM concentration (5 ppm). In order to minimize the effect of instrumental variability, all analyses were completed in one working day. 71 72 73 74 75 76 DI-ESI-HRMS analysis settings Prior to the direct infusion experiment, all samples were treated as previously described and continuously loaded into the ESI source by a Hamilton syringe (5 µl) using the instrument s built-in syringe pusher, operating at 8µL min -1. Syringe and ion source were connected by PEEK tubing. To avoid any sample carry-over, the infusion equipment was rinsed with 5:5 ultrapure acetonitrile/water solution between samples. S3
77 78 79 In order to compare the obtained results to the chromatographic experiment, the same ESI and Orbitrap settings were maintained during the direct infusion analysis. spectra were acquired for each sample. Sample code 81 82 83 84 85 86 87 88 89 To facilitate the interpretation of Bray-Curtis dissimilarity and Pearson correlation coefficients tables (Tables.xlsx, sheets names: Bray-Curtis dissimilarity and Pearson corr. coefficients) a sample code was established. For example, the result obtained by the first HPLC injection of the marine sample, was named M-lc1_: M identifies the water source, lc stands for liquid chromatography, while the replicate is identified by the first number and the segments ( 1-7 ) the last number ( identify the result from the whole chromatographic run, previous to the in silico fractionation). Each sample was run in in triplicate. The initials Rand L identify the river and lake sample respectively. Results from DI approach were named similarly: capital letter (M, R and L) for source identification followed by di. S4
9 91 92 93 94 95 96 97 98 Figure S1 Above: chromatogram from SRFA sample (orange), the acetonitrile percentage (ACN%) employed in the step gradient is showed in blue; below: chromatogram from the blank injection (black), performed with a faster gradient (blue) immediately after the injection of the SRFA sample, in order to evaluate the sample carry-over. The dashed lines (in red) represent the gradient shift due to the system dead volume. Table S1 Elution settings of the three step gradient employed for the HPLC analysis of the DOM samples. Time (min) Flow rate (µl min -1 ) % H 2 O +.1% formic acid % Acetonitrile 99 95 5 2 2.1 5 12 5 13 5 22 5 23 5 1 9 32 5 1 9 33 5 95 5 45 5 95 5 S5
11 12 13 14 Figure S2 a) Left: van Krevelen diagram showing all assigned peaks in the marine sample by the LC method (blue, 3449 formulas), and DI method (grey, 335 formulas); right: method-specific assignments: 877 LC-formulas (blue) and 463 DI-formulas (grey); b) corresponding Kendrick diagrams of data described in van Krevelen areas in panel a. 15 16 17 18 19 Figure S3 a) Left: van Krevelen diagram showing all assigned peaks in the lake sample by the LC method (green, 4979 formulas), and DI method (grey, 64 formulas); right: method-specific assignments: 1264 LC-formulas (green) and 349 DI-formulas (grey); b) corresponding Kendrick diagrams of data described in van Krevelen areas in panel a. S6
11 Y:\Jeff\claSIfig\claSI_1uLb 8/9/17 14:28:9 Y:\Jeff\claSIfig\claSI_1uLb 8/9/17 14:28:9 111 112 113 114 RT:.1-44.97 Relative Abundance 8.77 9.1 9.45 19.9 18.77 19.55.24 7.33 12.78 21. 5.69 27.27 27.76 39.33.24 33.64 8.47 8.84 8.98 1.8 12.9 19..6. 2.23 22. 24.62.34 32.97 2.45 8.64 7.81 7.3 7.98 8. 9.38 19.12 1.31 12.64.71 8.54 9.29 19.29 9.69 1.27 6.92 11.68 3. 14.11 8.67 8.97 8.36 19.28 9.74 7.3 1.43 19.86 21.51 3.96 13.59.55 5 1 15 25 3 35 Time (min) 23.99 27.36 31.97 39.77 22.11 24.29 28.7 34.71 38.63.63 21.26 1.79 23.97 26.13 3.15 38.72.64 1.1E7 TIC MS clasi_1ulb 2.64E7 TIC MS clasi_5ul 8.8E7 TIC MS clasi_ul 1.53E8 TIC MS clasi_5ul 2.33E8 TIC MS clasi_ul RT:.1-44.97 Relative Abundance 5 1 15 25 3 35 Figure S4 Chromatograms demonstrating the reproducibility of the method at different loading volumes from 1µL (top), 5,,5 µl, to µl (bottom). Left panel is the total ion current, right panel shows mass 369.8 with the three model compounds added at 3ppb in 5ppm SRFA. 9.23 9.48 9.75 8.49 12.62 7.2 13.15 18.98 19.52 5.69 23.81 27.82 32.13 38.89 41.45 7.44 8.14 9.22 9.68 12.68 6.49 13.1 19. 5.87 8.21 7.52 7.4 9.19 9.67 12.5 19.57 Time (min) 23.91 33.4 38.66 42.57 19.74 19.12 5.53.65 24.75 39.1 32.77 43.48 7.56 9.18 7.11 6.55 8.54 9.39 12.37 13.14.11 34. 5.5.65 26.37 38.57 29.5 41.2 7.27 7.9 9.23 12.42 13.6.1 6.68 3.96 34.52.52 24.83 38.48 3.46 29.58.36 3.9E4 7.89E4 3.E5 6.77E5 1.27E6 m/z= 369.8-369.9 MS clasi_1ulb m/z= 369.8-369.9 MS clasi_5ul m/z= 369.8-369.9 MS clasi_ul m/z= 369.8-369.9 MS clasi_5ul m/z= 369.8-369.9 MS clasi_ul 115 S7
116 117 118 119 1 121 Figure S5 Comparison of the length of CH 2 homologous series (Kendrick series) in the marine sample, segment 2 compared with segment 6. The 6 th segment has homologous series extending to 19 members, compared with only 13 for segment 2. For this analysis, we did not check for entirely continuous series, but rather the number of formulas with the same number of oxygen ( m ), unsaturation (z) and nitrogen ( p ) in a formula C n H (2n-z) O m N p 122 % N-containing formulas 5 3 1 1 2 3 4 5 6 7 Segment SRFA Sea Lake 123 124 125 Figure S6 The histograms express the percentage of N-containing formulas relative to the total assignments of the considered segment. The color code indicates the three water sources analyzed: SRFA (orange), marine (blue) and lake (green). S8
126 127 128 Figure S7 van Krevelen diagrams of N-containing compounds. Marine (blue) and Lake (green): segments from 1 to 7. Shared compounds are in displayed in magenta. 129 S9
13 131 132 133 134 135 136 REFERENCES (1) Green, N. W.; Perdue, E. M.; Aiken, G. R.; Butler, K. D.; Chen, H.; Dittmar, T.; Niggemann, J.; Stubbins, A. An intercomparison of three methods for the large-scale isolation of oceanic dissolved organic matter. Mar. Chem. 14, 161, 14 19. (2) Hawkes, J. A.; Dittmar, T.; Patriarca, C.; Tranvik, L. J.; Bergquist, J. Evaluation of the Orbitrap mass spectrometer for the molecular fingerprinting analysis of natural dissolved organic matter (DOM). Anal. Chem. 16, 88 (15), 7698 774. 137 S1