TY - JOUR
AU - Höffler, Susanne
AB - Abstract An high performance liquid chromatography (HPLC) method, applying diode array (DAD) and coulometric array detection simultaneously, was developed to determine monomeric phenolic compounds in lignin hydrolysates. To allow for additional mass spectrometric detection, a formic acid gradient was designed as an alternative to the non-volatile phosphate buffer gradient. Within a total run time of 28 min, 10 out of 14 analytes were baseline separated by the formic acid gradient, and 8 compounds were completely resolved by the phosphate gradient. The dependence of the coulometric detection on cell potential, potential step width, eluent pH and eluent composition was comprehensively tested. Detection limits ≤20 μg/L were achieved by DAD detection for nine analytes and by coulometric array detection for all analytes, demonstrating the superior sensitivity of electrochemical detection. DAD offered better linearity and reproducibility. Inter-day precision of peak heights spanned from 0.26 to 2.00% (formic acid gradient) and from 0.75 to 2.85% (phosphate buffer gradient) operating the DAD detector, but it exceeded 10% for several compounds applying the coulometric array detector. The simultaneous operation of both detectors offered an enhanced certainty of substance identification. The appropriateness of the method was confirmed by the analysis of various hydrolysates of lignin-containing materials. Introduction Lignin is one of the most important biopolymeric materials on earth, surpassed (in terms of mass) by cellulose only. It is a constituent of all woody plants and reaches mass proportions up to 30% in coniferous trees. Its biosynthesis starts with the aromatic amino acid phenylalanine via the formation of the three main building blocks, coniferyl, sinapyl and p-coumaryl alcohol, all belonging to the generic group of p-phenylpropenoic alcohols (1). Due to the multitude of reactions transforming the monomeric units into the copolymer and due to the diversity of the hereby formed bonding structures, lignin is a highly heterogeneous and variably composed material. Currently, in most technical processes exploiting lignocellulosic biomass, e.g., pulping and paper production, lignin is treated as a waste (2). Conventional technologies to remove lignin, e.g., kraft pulping, generate huge amounts of concentrated and toxic wastewaters, often leading to a serious pollution of the receiving water bodies (3, 4). With the increasing awareness of the necessity to improve the degree of sustainability of chemical products and processes and to replace the fossil resources by renewables, the physical and chemical properties of lignin are re-evaluated. In some cases, lignin is directly used in its original form as heteropolymer, e.g., for soil amendment, activated carbon production or as filter material in polymer manufacturing (5). Newer approaches focus on lignin as a potential source for value chemicals like phenols since this biopolymer represents the largest renewable resource of aromatic compounds. Typical monomeric products of the chemical depolymerization of lignin are, among others, vanillin, vanillic acid, acetovanillone and acetosyringone (5). The qualitative and quantitative composition of the conversion products depend on many factors, e.g., raw material, pretreatment, type of transformation reaction and reaction conditions (6–9). With the intensification of the material valorization of lignin, the need for fast, robust and precise analytical methods for the determination of monomeric products of its hydrolysis and conversion, i.e., phenolic compounds and related derivatives, is growing. Mixtures of these compounds are usually separated by liquid chromatography (LC), gas chromatography (GC) or capillary electrophoresis (CE) in combination with UV-VIS, diode array (DAD), fluorimetric, electrochemical or mass spectrometric (MS) detection (10–12). High performance liquid chromatography (HPLC) is the preferred separation technique for aqueous solutions of phenolic compounds. Mainly reversed phase (RP) separation columns were selected for this task, but other stationary phases, enabling ion exchange or ion exclusion processes, were applied too, albeit with a lower overall separation efficiency (13, 14). In conjunction with RP–HPLC separation, lignin monomers are detected by UV absorption (15), DAD (12, 16) or MS/MS (17). Several investigations describe the application of various types of electrochemical devices for the determination of phenolic compounds in (processed) plants, fruits and beverages (18, 19), but these detectors are very seldom used for lignin hydrolysate analysis. To the best of our knowledge, no method is available describing the application of the most powerful electrochemical detector, the coulometric electrode array (CoulArray) detector, for that purpose. This device differs from the more convenient amperometric detector in the almost complete oxidative or reductive electron transfer from the analyte to the electrode surface (superior sensitivity) and in the simultaneous registration of signals from multiple (up to 16) electrodes, working at different potentials (multichannel or array detection), providing superior selectivity (20, 21). Since coulometric array detection provides analytical information complementary to photodiode array detection (DAD), the combination of both is expected to further enhance the selectivity and validity of analytical results. Thus, it was the intention of this study to elaborate a RP-HPLC method for the analysis of lignin hydrolysates by applying a DAD and a CoulArray detector in series. Furthermore, the sensitivity of the detectors should be compared, and the overall performance of the method should be checked by the investigation of lignin degradation products. Experimental Chemicals and materials The following chemicals stemmed from Sigma-Aldrich (Darmstadt, Germany): 3-hydroxybenzaldehyde (purity 97%), 4-hydroxybenzaldehyde (98%), p-coumaric acid (98%), acetosyringone (97%) and syringic acid (>95%). Vanillic acid (>98%), ferulic acid (98%), vanillin (>99%) and syringaldehyde (>98%) were from Carl Roth (Karlsruhe, Germany), as well as HPLC-MS grade methanol. 3,4-dihydroxybenzoic acid (protocatechuic acid, 97%), iso-vanillic acid (>97%) and acetovanillone (98%) were purchased from Alfa Aesar (Karlsruhe, Germany). Salicylic acid (puriss.), formic acid (98–100%), NaH2PO4 × H2O (puriss.) and (NH4)2Fe (SO4)2 × 6 H2O were obtained from Merck (Darmstadt, Germany). Benzoic acid (>99.7%) and acetonitrile (ACN, HPLC grade) were delivered by VWR International (Darmstadt, Germany). Alkaline lignin was procured from TCI (Eschborn, Germany), and copper (II) oxide was stemmed from SCS GmbH (Bonn, Germany). Deionized water was prepared with a Nanopure Astacus water filtration system (Membrapure, Berlin, Germany). Sawdust and various wood specimens were received from the joinery of Trier University. Digestion of lignin-containing materials To depolymerize lignin and to yield phenolic monomers, alkaline lignin and lignocellulosic materials were digested under oxidizing conditions. 200 mg of the shredded particles (<1 mm) were filled together with 110 mg of CuO, and 35 mg of (NH4)2Fe (SO4)2 and 10 ml of 2M NaOH into screw-top teflon vessels. The filled vessels were mounted in the sample carousel of a microwave digestion device (Ethos, MLS GmbH, Leutkirch, Germany). Applying a temperature step gradient (maximum temperature: 150°C, reached after approximately 18 min), the samples were digested during 20 min. The cooled reaction mixtures were passed through folded filters. Then the volumes of the filtrates were adjusted to 100 ml by the addition of distilled water. For HPLC analysis, aliquota of these solutions were further diluted by addition of the 4-fold volume of the chromatographic eluent (start composition of the gradient). Instrumental Analytical method development was conducted with an ICS-5000 LC system (ThermoFisher Scientific, Dreieich, Germany), which was equipped with an AS-AP autosampler, a photodiode array detector (PDA) and an eight-channel coulometric detector CoulArray® 5600A (ESA, Chelmsford, USA). Data acquisition and processing were handled by a Chromeleon 7 workstation and by the built-in CoulArray® software. The analytes were separated on a Kinetex C18 RP column (100 × 3 mm, particle size 2.6 μm, Phenomenex, Aschaffenburg, Germany), which was equipped with a corresponding guard column. Injection volume was 10 μL. All samples were membrane-filtered (0.2 μm) prior to injection. Various phosphate buffers and aqueous formic acid solutions served as the mobile phase in an isocratic and gradient elution. Specific chromatographic conditions, e.g., eluent composition, pH value, gradient profile, temperature, flow rate and detector settings, were subject of the method development. Determination of analytical quality parameters The linear ranges and the corresponding calibration functions were deduced from the measurements of dilution series including up to 11 concentrations of the respective stock solution. Calibrations were performed, based on peak heights, for each gradient profile and each detector specifically. Processed peak heights are the means of three subsequent sample injections. A signal-to-noise (S/N) ratio of 3 served as the lowest detection limit (LOD), and an S/N ratio of 10 defined the lowest quantitation limit (LOQ). The intra-day and inter-day precision of the various methodical variants was evaluated by multiple injections of a mix standard, containing 1 mg/L of every analyte. The intra-day precision was deduced from nine injections. Data from three injections per day over 6 days served for the calculation of the inter-day precision. The analyte recoveries from the conversion of lignocellulosic materials and lignin were established with and without the integration of the digestion process by spiking known amounts of standards into the test samples. To check the recoveries for the whole process, the analytes (amounts between 1.0 and 8.0 mg) were added to the reaction mixtures prior to the microwave treatment. Alternatively, analyte solutions were spiked to the hydrolysates generated by microwave digestion, leading to concentration increases between 1.0 and 5.0 mg/L. Results Optimization of the chromatographic conditions A few studies describing the separation of phenolic compounds on RP columns, followed by CoulArray detection, served as starting points for method development (22, 23). Various phosphate buffers were screened as eluents. Since acidic compounds are sufficiently retained on RP columns only if mainly present in their non-dissociated form, a buffer pH value below the pKa values of the analytes is required. Due to the relatively high pKa values of phenolic aldehydes and ketones, pH changes in the acidic range might not have a significant influence on their retention behavior. Best separation results were achieved with the following linear gradient, where component A was a 0.1 M aqueous solution of NaH2PO4 (pH 3.35), and component B contained 60% (v/v) of ACN, 30% (v/v) of 0.03 M aqueous NaH2PO4 solution (pH 3.45) and 10% (v/v) of methanol. The steps were (i) 0–10 min: linear gradient from 100 to 94% A, (ii) 10–34 min: linear gradient from 94 to 77.5% A and (iii) 34–39 min: linear gradient from 77.5 to 0% A. Flow rate was 0.4 ml/min and temperature was 39°C (Figure 1). As Figure 1 illustrates, mostly baseline separation was achieved but not for the substance pairs, 4-hydroxybenzaldehyde /vanillic acid, 3-hydroxybenzaldehyde/iso-vanillic acid and syringaldehyde/acetovanillone. Due to the low volatility of phosphate salts, the eluent is not suited for MS detection. To provide an MS-compatible alternative, a second chromatographic method was elaborated, replacing phosphate by formic acid. Various formic acid concentrations constituted in water and methanol were tested. Figure 1 Open in new tabDownload slide Separation of 14 phenolic reference compounds with optimized phosphate buffer gradient (pH: 3.35); peaks: 1: 3,4-dihydroxybenzoic acid, 2: 4-hydroxybenzaldehyde, 3: vanillic acid, 4: salicylic acid, 5: 3-hydroxybenzaldehyde, 6: iso-vanillic acid, 7: syringic acid, 8: vanillin, 9: p-coumaric acid, 10: benzoic acid, 11: syringaldehyde, 12: acetovanillone, 13: ferulic acid and 14: acetosyringone; analyte concentrations: 10 mg/L, DAD wavelength: 216 nm. Figure 1 Open in new tabDownload slide Separation of 14 phenolic reference compounds with optimized phosphate buffer gradient (pH: 3.35); peaks: 1: 3,4-dihydroxybenzoic acid, 2: 4-hydroxybenzaldehyde, 3: vanillic acid, 4: salicylic acid, 5: 3-hydroxybenzaldehyde, 6: iso-vanillic acid, 7: syringic acid, 8: vanillin, 9: p-coumaric acid, 10: benzoic acid, 11: syringaldehyde, 12: acetovanillone, 13: ferulic acid and 14: acetosyringone; analyte concentrations: 10 mg/L, DAD wavelength: 216 nm. Different from the phosphate buffer gradient, analyte retention decreased with decreasing pH, i.e., increasing the formic acid concentration. Since the pH was ≤2.55 with all the tested gradient profiles, this effect might not be provoked by alterations of the dissociation degrees of the analytes, except salicylic acid. Presumably, a faster analyte displacement with an increasing number of formic acid molecules was responsible for that phenomenon. Since resolution between 3-hydroxybenzaldehyde and vanillic acid decreased with increasing formic acid concentration, whereas that of the pair p-coumaric acid/syringaldehyde increased, the finally selected eluent pH of 2.27 offered a practicable compromise. The corresponding gradient composition was as follows: Component A was a 0.15 M aqueous formic acid solution (pH 2.27), and component B contained the same formic acid concentration, established in methanol. The gradient steps were (i) 0–20 min: isocratic: 90% A, 10% B, (ii) 20–35 min: linear gradient from 90 to 75% A and (iii) 35–40 min: linear gradient from 75 to 50% A. Flow rate was 0.4 ml/min and temperature was 40°C (Figure 2). Figure 2 Open in new tabDownload slide Separation of 14 phenolic reference compounds (10 mg/L each) with optimized formic acid gradient (pH: 2.27); peaks: 1: 3,4-dihydroxybenzoic acid, 2: 4-hydroxybenzaldehyde, 3: 3-hydroxybenzaldehyde, 4: vanillic acid, 5: iso-vanillic acid, 6: vanillin, 7: syringic acid, 8: p-coumaric acid, 9: syringaldehyde, 10: acetovanillone, 11: benzoic acid, 12: salicylic acid, 13: ferulic acid and 14: acetosyringone; analyte concentrations: 10 mg/L, DAD wavelength: 231 nm. Figure 2 Open in new tabDownload slide Separation of 14 phenolic reference compounds (10 mg/L each) with optimized formic acid gradient (pH: 2.27); peaks: 1: 3,4-dihydroxybenzoic acid, 2: 4-hydroxybenzaldehyde, 3: 3-hydroxybenzaldehyde, 4: vanillic acid, 5: iso-vanillic acid, 6: vanillin, 7: syringic acid, 8: p-coumaric acid, 9: syringaldehyde, 10: acetovanillone, 11: benzoic acid, 12: salicylic acid, 13: ferulic acid and 14: acetosyringone; analyte concentrations: 10 mg/L, DAD wavelength: 231 nm. The noticeable baseline increase starting from about 23 min is caused by the increasing methanol content of the eluent, reducing its optical transparency at a detection wavelength of 231 nm. Applying DAD, higher wavelengths, e.g., 258, 275 and 305 nm, respectively, were selected for simultaneous absorbance recording, but none offered a comparable, balanced sensitivity for all analytes. Due to the essentially lower maximal methanol content of the phosphate gradient, detection at lower wavelengths is possible without an elevated background absorbance. Chromatographic selectivity is similar to the phosphate buffer gradient except the much higher retention of salicylic acid due to the lower eluent pH and subsequent lower dissociation of this acid, which had the highest acidity (pKa: 2.75) of all analytes. The predominantly lower retention might be attributed to the higher displacement efficiency of the formic acid molecules. As already stated, complete baseline resolution was also not possible with this eluent, but the affected substance pairs did not match with those generated by the phosphate buffer gradient. Optimization of the CoulArray operation conditions Generally, coulometric detection is based on the quantification of the current amount which is produced or consumed by the oxidation or reduction of the analyte at a given electrode potential. The current is highest, applying a working potential close to the conditional (at given reaction conditions) redox potential of the analyte, i.e., the half-wave potential. A coulometric detector differs from an amperometric one in the design of the electrode and in the corresponding eluent flow. The coulometric detector houses a porous electrode allowing for a flow through of the eluent, whereas the electrode of an amperometric detector is non-porous with a comparably small surface area. Due to the essentially higher active surface area of the coulometric electrode, the electrochemical conversion of the analytes might approach 100% under ideal conditions, offering a considerable increase in the detection sensitivity. Due to the array arrangement of the coulometric electrode cells, the effluent of the separation column passes successively through eight miniaturized cells (“channels”) with stepwise increasing potentials. The analytical information comprises the current amounts registered at every potential. The highest current amount (peak area) or current flow (peak height) is selected for quantification and the channel, where the corresponding potential is applied, is termed the “main” or “dominant” channel. Thus, to achieve a high detection sensitivity, most of the electrochemical analyte conversion should take place at this potential. To realize those conditions, often comparably high potential differences between the channels are required. On the other hand, these differences should not exceed a certain size since otherwise the qualitative analytical information (correlation between the molecular structure and main electrochemical potential) offered by the array arrangement would get lost. Consequently, detection conditions, providing an optimal balance between both demands, have to be identified. Essential results of this procedure, applying the phosphate buffer gradient, are summarized in Table I. Depending on the lower limit (minimum potential) of the total potential range and the constant voltage difference between consecutive cells, different main channel potentials varied within a range of 150 mV or less. Working potentials between 500 and 800 mV were best suited for the majority of the compounds. Due to the presence of two vicinal hydroxyl groups (catechol-like structure), 3,4-dihydroxybenzoic acid is oxidized at a significantly lower potential (250 mV). Also, syringaldehyde is oxidizable at somewhat lower potential (400 mV). Due to its missing oxidizability, it was not feasible to coulometrically determine benzoic acid under tested conditions. In most cases, alterations of the minimal potential and potential differences were not paralleled by large changes of the detection sensitivity. Since conditions were searched enabling sensitive and reproducible detection of most of the analytes, a minimal potential of 0 mV and a potential difference of 150 mV were found to be optimal, except for 4-hydroxybenzaldehyde and acetovanillone. Table I Dependence of the CoulArray Peak Heights and of the Main Channel Potential on Selected Minimal Potential and Potential Differences between the Array Channels Analytes . Minimal potential (mV) . Potential difference (mV) . Main channel potential (mV) . Peak height (μA) . 3,4-Dihydroxybenzoic acid 200 100 300 6.88   0 150 300 6.35   100 150 250 6.58 4-Hydroxybenzaldehyde 0 125 750 5.55   400 200 800 5.63   0 150 750 4.20 Vanillic acid 0 150 600 4.08   300 200 500 3.29   65 125 565 3.60 iso-Vanillic acid 0 150 600 3.71 Salicylic acid 400 200 800 6.49 3-Hydroxybenzaldehyde 0 150 800 6.06   0 125 750 5.82 Syringic acid 0 150 600 3.17 Vanillin 0 150 600 4.33   100 150 700 3.84   65 125 565 3.97 p-Coumaric acid 0 150 600 4.33   65 125 565 3.63   100 150 700 3.74 Syringaldehyde 100 150 400 3.16   0 150 600 3.09   300 200 500 3.05   200 100 400 3.02 Acetovanillone 50 150 650 3.09   100 150 700 2.10   400 200 600 2.68   0 150 750 2.17 Ferulic acid 0 150 600 2.82   300 200 500 2.79 Acetosyringone 0 150 600 2.78   300 200 500 2.63 Analytes . Minimal potential (mV) . Potential difference (mV) . Main channel potential (mV) . Peak height (μA) . 3,4-Dihydroxybenzoic acid 200 100 300 6.88   0 150 300 6.35   100 150 250 6.58 4-Hydroxybenzaldehyde 0 125 750 5.55   400 200 800 5.63   0 150 750 4.20 Vanillic acid 0 150 600 4.08   300 200 500 3.29   65 125 565 3.60 iso-Vanillic acid 0 150 600 3.71 Salicylic acid 400 200 800 6.49 3-Hydroxybenzaldehyde 0 150 800 6.06   0 125 750 5.82 Syringic acid 0 150 600 3.17 Vanillin 0 150 600 4.33   100 150 700 3.84   65 125 565 3.97 p-Coumaric acid 0 150 600 4.33   65 125 565 3.63   100 150 700 3.74 Syringaldehyde 100 150 400 3.16   0 150 600 3.09   300 200 500 3.05   200 100 400 3.02 Acetovanillone 50 150 650 3.09   100 150 700 2.10   400 200 600 2.68   0 150 750 2.17 Ferulic acid 0 150 600 2.82   300 200 500 2.79 Acetosyringone 0 150 600 2.78   300 200 500 2.63 Phosphate buffer gradient, means of three injections, injection volume: 10 μL and analyte concentrations: 5 mg/L. In italics: conditions with low signal reproducibility. Open in new tab Table I Dependence of the CoulArray Peak Heights and of the Main Channel Potential on Selected Minimal Potential and Potential Differences between the Array Channels Analytes . Minimal potential (mV) . Potential difference (mV) . Main channel potential (mV) . Peak height (μA) . 3,4-Dihydroxybenzoic acid 200 100 300 6.88   0 150 300 6.35   100 150 250 6.58 4-Hydroxybenzaldehyde 0 125 750 5.55   400 200 800 5.63   0 150 750 4.20 Vanillic acid 0 150 600 4.08   300 200 500 3.29   65 125 565 3.60 iso-Vanillic acid 0 150 600 3.71 Salicylic acid 400 200 800 6.49 3-Hydroxybenzaldehyde 0 150 800 6.06   0 125 750 5.82 Syringic acid 0 150 600 3.17 Vanillin 0 150 600 4.33   100 150 700 3.84   65 125 565 3.97 p-Coumaric acid 0 150 600 4.33   65 125 565 3.63   100 150 700 3.74 Syringaldehyde 100 150 400 3.16   0 150 600 3.09   300 200 500 3.05   200 100 400 3.02 Acetovanillone 50 150 650 3.09   100 150 700 2.10   400 200 600 2.68   0 150 750 2.17 Ferulic acid 0 150 600 2.82   300 200 500 2.79 Acetosyringone 0 150 600 2.78   300 200 500 2.63 Analytes . Minimal potential (mV) . Potential difference (mV) . Main channel potential (mV) . Peak height (μA) . 3,4-Dihydroxybenzoic acid 200 100 300 6.88   0 150 300 6.35   100 150 250 6.58 4-Hydroxybenzaldehyde 0 125 750 5.55   400 200 800 5.63   0 150 750 4.20 Vanillic acid 0 150 600 4.08   300 200 500 3.29   65 125 565 3.60 iso-Vanillic acid 0 150 600 3.71 Salicylic acid 400 200 800 6.49 3-Hydroxybenzaldehyde 0 150 800 6.06   0 125 750 5.82 Syringic acid 0 150 600 3.17 Vanillin 0 150 600 4.33   100 150 700 3.84   65 125 565 3.97 p-Coumaric acid 0 150 600 4.33   65 125 565 3.63   100 150 700 3.74 Syringaldehyde 100 150 400 3.16   0 150 600 3.09   300 200 500 3.05   200 100 400 3.02 Acetovanillone 50 150 650 3.09   100 150 700 2.10   400 200 600 2.68   0 150 750 2.17 Ferulic acid 0 150 600 2.82   300 200 500 2.79 Acetosyringone 0 150 600 2.78   300 200 500 2.63 Phosphate buffer gradient, means of three injections, injection volume: 10 μL and analyte concentrations: 5 mg/L. In italics: conditions with low signal reproducibility. Open in new tab The eluent-pH does not only influence chromatographic separation but also the electrochemical detection sensitivity. CoulArray detection sensitivities were compared for two formic acid eluent concentrations and the respective pH values. Both eluents were tested at two detector operation conditions, differing in the total potential range and voltage differences of the working potentials of the individual electrodes. As Table II reveals, detection at smaller potential and potential difference is more sensitive for the majority of compounds at both pH values, except for the hydroxybenzaldehydes, p-coumaric acid and salicylic acid. In almost all cases, sensitivity at pH 2.27 was superior to pH 2.20. Figure 3 presents an RP-CoulArray chromatogram of a reference compound standard, applying the formic acid (0.15 mol/L, pH 2.27) eluent. Most analytes were detected at nearly every cell potential, but the respective degrees of the oxidative substance conversions were very different, leading to the definition of three dominant channels at 850 , 900 and 1050 mV, respectively. Table II Dependence of CoulArray Peak Heights (μA) on Selected Potential Range, Step Width and pH Value of the Formic Acid Eluent pH value . 2.20 . 2.27 . Potential range (mV) 550–1250 850–1200 550–1250 850–1200 Step width (mV) 100 50 100 50 3,4-Dihydroxybenzoic acid 9.51 13.2 8.40 13.1 4-Hydroxybenzaldehyde 7.01 5.73 7.12 5.86 3-Hydroxybenzaldehyde 7.38 4.09 7.83 4.31 Vanillic acid 5.86 6.31 6.31 6.91 iso-Vanillic acid 4.60 5.91 5.22 6.64 Vanillin 5.24 4.57 5.79 5.26 Syringic acid 3.37 4.28 3.51 5.20 p-Coumaric acid 2.75 1.97 3.66 2.21 Syringaldehyde 2.52 2.77 2.57 3.15 Acetovanillone 2.67 2.68 3.09 3.43 Salicylic acid 3.30 1.66 3.80 1.87 Ferulic acid 1.79 2.92 1.85 3.87 Acetosyringone 2.16 2.26 2.35 3.63 pH value . 2.20 . 2.27 . Potential range (mV) 550–1250 850–1200 550–1250 850–1200 Step width (mV) 100 50 100 50 3,4-Dihydroxybenzoic acid 9.51 13.2 8.40 13.1 4-Hydroxybenzaldehyde 7.01 5.73 7.12 5.86 3-Hydroxybenzaldehyde 7.38 4.09 7.83 4.31 Vanillic acid 5.86 6.31 6.31 6.91 iso-Vanillic acid 4.60 5.91 5.22 6.64 Vanillin 5.24 4.57 5.79 5.26 Syringic acid 3.37 4.28 3.51 5.20 p-Coumaric acid 2.75 1.97 3.66 2.21 Syringaldehyde 2.52 2.77 2.57 3.15 Acetovanillone 2.67 2.68 3.09 3.43 Salicylic acid 3.30 1.66 3.80 1.87 Ferulic acid 1.79 2.92 1.85 3.87 Acetosyringone 2.16 2.26 2.35 3.63 Means of three injections, analyte concentrations: 10 mg/L. Open in new tab Table II Dependence of CoulArray Peak Heights (μA) on Selected Potential Range, Step Width and pH Value of the Formic Acid Eluent pH value . 2.20 . 2.27 . Potential range (mV) 550–1250 850–1200 550–1250 850–1200 Step width (mV) 100 50 100 50 3,4-Dihydroxybenzoic acid 9.51 13.2 8.40 13.1 4-Hydroxybenzaldehyde 7.01 5.73 7.12 5.86 3-Hydroxybenzaldehyde 7.38 4.09 7.83 4.31 Vanillic acid 5.86 6.31 6.31 6.91 iso-Vanillic acid 4.60 5.91 5.22 6.64 Vanillin 5.24 4.57 5.79 5.26 Syringic acid 3.37 4.28 3.51 5.20 p-Coumaric acid 2.75 1.97 3.66 2.21 Syringaldehyde 2.52 2.77 2.57 3.15 Acetovanillone 2.67 2.68 3.09 3.43 Salicylic acid 3.30 1.66 3.80 1.87 Ferulic acid 1.79 2.92 1.85 3.87 Acetosyringone 2.16 2.26 2.35 3.63 pH value . 2.20 . 2.27 . Potential range (mV) 550–1250 850–1200 550–1250 850–1200 Step width (mV) 100 50 100 50 3,4-Dihydroxybenzoic acid 9.51 13.2 8.40 13.1 4-Hydroxybenzaldehyde 7.01 5.73 7.12 5.86 3-Hydroxybenzaldehyde 7.38 4.09 7.83 4.31 Vanillic acid 5.86 6.31 6.31 6.91 iso-Vanillic acid 4.60 5.91 5.22 6.64 Vanillin 5.24 4.57 5.79 5.26 Syringic acid 3.37 4.28 3.51 5.20 p-Coumaric acid 2.75 1.97 3.66 2.21 Syringaldehyde 2.52 2.77 2.57 3.15 Acetovanillone 2.67 2.68 3.09 3.43 Salicylic acid 3.30 1.66 3.80 1.87 Ferulic acid 1.79 2.92 1.85 3.87 Acetosyringone 2.16 2.26 2.35 3.63 Means of three injections, analyte concentrations: 10 mg/L. Open in new tab Figure 3 Open in new tabDownload slide CoulArray detection of phenolic reference compounds separated by the formic acid gradient (pH: 2.27); potential range: 850–1200 mV, step width: 50 mV; peaks: 1: 3,4-dihydroxybenzoic acid, 2: 4-hydroxybenzaldehyde, 3: 3-hydroxybenzaldehyde, 4: vanillic acid, 5: iso-vanillic acid, 6: vanillin, 7: syringic acid, 8: p-coumaric acid, 9: syringaldehyde, 10: acetovanillone, 11: salicylic acid, 12: ferulic acid and 13: acetosyringone; analyte concentrations: 10 mg/L. Figure 3 Open in new tabDownload slide CoulArray detection of phenolic reference compounds separated by the formic acid gradient (pH: 2.27); potential range: 850–1200 mV, step width: 50 mV; peaks: 1: 3,4-dihydroxybenzoic acid, 2: 4-hydroxybenzaldehyde, 3: 3-hydroxybenzaldehyde, 4: vanillic acid, 5: iso-vanillic acid, 6: vanillin, 7: syringic acid, 8: p-coumaric acid, 9: syringaldehyde, 10: acetovanillone, 11: salicylic acid, 12: ferulic acid and 13: acetosyringone; analyte concentrations: 10 mg/L. Performance of the developed analytical methods Table III summarizes the essential quality parameters of the chromatographic separation of the phenolic compounds applying the phosphate buffer gradient. Calibrations were conducted for both detectors specifically. To check the linearity of the detector responses within a large concentration range, dilution series, containing between 8 and 11 concentration levels, were analyzed for every compound. Mostly, the linear range stretched at least over 2.5 orders of magnitude for both detectors. With respect to the coefficients of determination (r2), most of the calibration functions, derived from DAD data, are marked by a very high degree of linearity. In the case of the coulometric data, linearity is somewhat lower, but it is still on an acceptable level. DAD detection limits, calculated based on an S/N ratio of 3, were in the range between 10 and 30 μg/L typically. LODs ≤15 μg/L were attained for acetovanillone, 3,4-dihydroxybenzoic acid, 4-hydroxybenzaldehyde, salicylic acid and vanillin, respectively. The CoulArray detection limits were slightly lower, ranging from 7 to 18 μg/L for all compounds. Coulometric detection was approximately 8-fold more sensitive than DAD for acetosyringone. Table III Comparison of DAD and CoulArray Detection Performance (Peak Heights), Applying the Phosphate Buffer Gradient Analytes . DAD . CoulArray .   . RT (min) . λ (nm) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Main channel (mV) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Acetosyringone 27.17 298 y = 5.0361x + 0.0438 0.9999 250–7500 54 180 600 y = 0.5427x + 0.004 0.9991 25–7500 7 23 Acetovanillone 23.06 228 y = 6.5898x + 0.1314 0.9999 25–7500 15 50 600 y = 0.4761x + 0.0657 0.9976 25–7500 16 53 Benzoic acid 21.64 228 y = 6.2335x + 0.0696 1.0000 50–7500 19 63 n.d. n.d. n.d. n.d. n.d. n.d. p-Coumaric acid 20.42 307 y = 10.786x − 0.0331 1.0000 50–7500 18 60 600 y = 0.783x + 0.0103 0.9978 25–7500 12 40 3.4-Dihydroxy-benzoic acid 3.78 258 y = 13.902x + 0.0933 0.9998 25–7500 9 30 300 y = 1.2573x − 0.1127 0.9998 125–7500 18 60 Ferulic acid 25.75 321 y = 6.3897x − 0.05 0.9998 125–7500 39 130 600 y = 0.5121x − 0.0387 0.9985 50–7500 14 47 3-Hydroxy-benzaldehyde 12.57 255 y = 5.7885x − 0.0048 1.0000 50–7500 29 97 750 y = 0.7323x + 0.0275 0.9997 25–7500 12 40 4-Hydroxy-benzaldehyde 10.24 284 y = 8.8251x + 0.1108 0.9999 25–7500 14 47 750 y = 0.6519x + 0.0653 0.9992 25–7500 10 33 iso-Vanillic acid 13.12 258 y = 5.4719x + 0.0024 1.0000 125–7500 28 93 600 y = 0.5904x + 0.0048 0.9985 50–7500 15 50 Salicylic acid 12.00 231 y = 4.1151x − 0.0288 1.0000 50–7500 13 43 750 y = 0.8101x + 0.0823 0.9984 25–7500 7 23 Syringic acid 15.78 273 y = 4.6157x + 0.007 1.0000 50–7500 20 67 600 y = 0.5854x − 0.0126 0.9988 50–7500 15 50 Syringaldehyde 22.52 306 y = 7.7084x + 0.0515 1.0000 125–7500 35 117 600 y = 0.65x + 0.0287 0.9996 25–7500 11 37 Vanillic acid 10.85 259 y = 6.4539x + 0.0332 1.0000 50–7500 23 77 600 y = 0.7329x − 0.00007 0.9998 25–7500 9 30 Vanillin 17.01 230 y = 8.0657x + 0.1106 1.0000 50–7500 12 40 600 y = 0.7925x + 0.0324 0.9980 50–7500 15 50 Analytes . DAD . CoulArray .   . RT (min) . λ (nm) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Main channel (mV) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Acetosyringone 27.17 298 y = 5.0361x + 0.0438 0.9999 250–7500 54 180 600 y = 0.5427x + 0.004 0.9991 25–7500 7 23 Acetovanillone 23.06 228 y = 6.5898x + 0.1314 0.9999 25–7500 15 50 600 y = 0.4761x + 0.0657 0.9976 25–7500 16 53 Benzoic acid 21.64 228 y = 6.2335x + 0.0696 1.0000 50–7500 19 63 n.d. n.d. n.d. n.d. n.d. n.d. p-Coumaric acid 20.42 307 y = 10.786x − 0.0331 1.0000 50–7500 18 60 600 y = 0.783x + 0.0103 0.9978 25–7500 12 40 3.4-Dihydroxy-benzoic acid 3.78 258 y = 13.902x + 0.0933 0.9998 25–7500 9 30 300 y = 1.2573x − 0.1127 0.9998 125–7500 18 60 Ferulic acid 25.75 321 y = 6.3897x − 0.05 0.9998 125–7500 39 130 600 y = 0.5121x − 0.0387 0.9985 50–7500 14 47 3-Hydroxy-benzaldehyde 12.57 255 y = 5.7885x − 0.0048 1.0000 50–7500 29 97 750 y = 0.7323x + 0.0275 0.9997 25–7500 12 40 4-Hydroxy-benzaldehyde 10.24 284 y = 8.8251x + 0.1108 0.9999 25–7500 14 47 750 y = 0.6519x + 0.0653 0.9992 25–7500 10 33 iso-Vanillic acid 13.12 258 y = 5.4719x + 0.0024 1.0000 125–7500 28 93 600 y = 0.5904x + 0.0048 0.9985 50–7500 15 50 Salicylic acid 12.00 231 y = 4.1151x − 0.0288 1.0000 50–7500 13 43 750 y = 0.8101x + 0.0823 0.9984 25–7500 7 23 Syringic acid 15.78 273 y = 4.6157x + 0.007 1.0000 50–7500 20 67 600 y = 0.5854x − 0.0126 0.9988 50–7500 15 50 Syringaldehyde 22.52 306 y = 7.7084x + 0.0515 1.0000 125–7500 35 117 600 y = 0.65x + 0.0287 0.9996 25–7500 11 37 Vanillic acid 10.85 259 y = 6.4539x + 0.0332 1.0000 50–7500 23 77 600 y = 0.7329x − 0.00007 0.9998 25–7500 9 30 Vanillin 17.01 230 y = 8.0657x + 0.1106 1.0000 50–7500 12 40 600 y = 0.7925x + 0.0324 0.9980 50–7500 15 50 n.d., not detectable. aLimit of detection (S/N: 3:1); blimit of quantitation (S/N: 10:1). Open in new tab Table III Comparison of DAD and CoulArray Detection Performance (Peak Heights), Applying the Phosphate Buffer Gradient Analytes . DAD . CoulArray .   . RT (min) . λ (nm) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Main channel (mV) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Acetosyringone 27.17 298 y = 5.0361x + 0.0438 0.9999 250–7500 54 180 600 y = 0.5427x + 0.004 0.9991 25–7500 7 23 Acetovanillone 23.06 228 y = 6.5898x + 0.1314 0.9999 25–7500 15 50 600 y = 0.4761x + 0.0657 0.9976 25–7500 16 53 Benzoic acid 21.64 228 y = 6.2335x + 0.0696 1.0000 50–7500 19 63 n.d. n.d. n.d. n.d. n.d. n.d. p-Coumaric acid 20.42 307 y = 10.786x − 0.0331 1.0000 50–7500 18 60 600 y = 0.783x + 0.0103 0.9978 25–7500 12 40 3.4-Dihydroxy-benzoic acid 3.78 258 y = 13.902x + 0.0933 0.9998 25–7500 9 30 300 y = 1.2573x − 0.1127 0.9998 125–7500 18 60 Ferulic acid 25.75 321 y = 6.3897x − 0.05 0.9998 125–7500 39 130 600 y = 0.5121x − 0.0387 0.9985 50–7500 14 47 3-Hydroxy-benzaldehyde 12.57 255 y = 5.7885x − 0.0048 1.0000 50–7500 29 97 750 y = 0.7323x + 0.0275 0.9997 25–7500 12 40 4-Hydroxy-benzaldehyde 10.24 284 y = 8.8251x + 0.1108 0.9999 25–7500 14 47 750 y = 0.6519x + 0.0653 0.9992 25–7500 10 33 iso-Vanillic acid 13.12 258 y = 5.4719x + 0.0024 1.0000 125–7500 28 93 600 y = 0.5904x + 0.0048 0.9985 50–7500 15 50 Salicylic acid 12.00 231 y = 4.1151x − 0.0288 1.0000 50–7500 13 43 750 y = 0.8101x + 0.0823 0.9984 25–7500 7 23 Syringic acid 15.78 273 y = 4.6157x + 0.007 1.0000 50–7500 20 67 600 y = 0.5854x − 0.0126 0.9988 50–7500 15 50 Syringaldehyde 22.52 306 y = 7.7084x + 0.0515 1.0000 125–7500 35 117 600 y = 0.65x + 0.0287 0.9996 25–7500 11 37 Vanillic acid 10.85 259 y = 6.4539x + 0.0332 1.0000 50–7500 23 77 600 y = 0.7329x − 0.00007 0.9998 25–7500 9 30 Vanillin 17.01 230 y = 8.0657x + 0.1106 1.0000 50–7500 12 40 600 y = 0.7925x + 0.0324 0.9980 50–7500 15 50 Analytes . DAD . CoulArray .   . RT (min) . λ (nm) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Main channel (mV) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Acetosyringone 27.17 298 y = 5.0361x + 0.0438 0.9999 250–7500 54 180 600 y = 0.5427x + 0.004 0.9991 25–7500 7 23 Acetovanillone 23.06 228 y = 6.5898x + 0.1314 0.9999 25–7500 15 50 600 y = 0.4761x + 0.0657 0.9976 25–7500 16 53 Benzoic acid 21.64 228 y = 6.2335x + 0.0696 1.0000 50–7500 19 63 n.d. n.d. n.d. n.d. n.d. n.d. p-Coumaric acid 20.42 307 y = 10.786x − 0.0331 1.0000 50–7500 18 60 600 y = 0.783x + 0.0103 0.9978 25–7500 12 40 3.4-Dihydroxy-benzoic acid 3.78 258 y = 13.902x + 0.0933 0.9998 25–7500 9 30 300 y = 1.2573x − 0.1127 0.9998 125–7500 18 60 Ferulic acid 25.75 321 y = 6.3897x − 0.05 0.9998 125–7500 39 130 600 y = 0.5121x − 0.0387 0.9985 50–7500 14 47 3-Hydroxy-benzaldehyde 12.57 255 y = 5.7885x − 0.0048 1.0000 50–7500 29 97 750 y = 0.7323x + 0.0275 0.9997 25–7500 12 40 4-Hydroxy-benzaldehyde 10.24 284 y = 8.8251x + 0.1108 0.9999 25–7500 14 47 750 y = 0.6519x + 0.0653 0.9992 25–7500 10 33 iso-Vanillic acid 13.12 258 y = 5.4719x + 0.0024 1.0000 125–7500 28 93 600 y = 0.5904x + 0.0048 0.9985 50–7500 15 50 Salicylic acid 12.00 231 y = 4.1151x − 0.0288 1.0000 50–7500 13 43 750 y = 0.8101x + 0.0823 0.9984 25–7500 7 23 Syringic acid 15.78 273 y = 4.6157x + 0.007 1.0000 50–7500 20 67 600 y = 0.5854x − 0.0126 0.9988 50–7500 15 50 Syringaldehyde 22.52 306 y = 7.7084x + 0.0515 1.0000 125–7500 35 117 600 y = 0.65x + 0.0287 0.9996 25–7500 11 37 Vanillic acid 10.85 259 y = 6.4539x + 0.0332 1.0000 50–7500 23 77 600 y = 0.7329x − 0.00007 0.9998 25–7500 9 30 Vanillin 17.01 230 y = 8.0657x + 0.1106 1.0000 50–7500 12 40 600 y = 0.7925x + 0.0324 0.9980 50–7500 15 50 n.d., not detectable. aLimit of detection (S/N: 3:1); blimit of quantitation (S/N: 10:1). Open in new tab Table IV lists the corresponding analytical quality parameters for the formic acid gradient. Baseline noise of this eluent is higher and optical transparency, especially in the low UV wavelength range, is lower than the phosphate gradient. This led to a decreased linearity of some calibration functions and to elevated detection limits in several cases. Compared with the phosphate gradient, strong reductions of the DAD detection sensitivity are noticeable for benzoic acid, 3-hydroxybenzaldehyde and salicylic acid. Except these compounds, limits of detection by DAD spanned from 7 to 43 μg/L. Table IV Comparison of DAD and CoulArray Detection Performance (Peak Heights), Applying the Formic Acid Gradient Analytes . DAD . CoulArray .   . RT (min) . λ (nm) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Main channel (mV) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Acetosyringone 26.73 298 y = 3.6395x + 0.1993 0.9994 125–10,000 43 143 850 y = 0.2586x + 0.0612 0.9941 50–10,000 24 80 Acetovanillone 18.55 275 y = 3.3964x + 0.1032 0.9998 50–10,000 29 97 900 y = 0.4124x + 0.054 0.9972 50–10,000 15 50 Benzoic acid 20.50 231 y = 4.4802x − 1.5685 0.999 2500–10,000 477 1590 n.d. n.d. n.d. n.d. n.d. n.d. p-Coumaric acid 16.25 308 y = 9.2165x − 0.0434 1.0000 50–10,000 14 47 900 y = 0.3892x + 0.1408 0.9875 25–10,000 15 50 3.4-Dihydroxy-benzoic acid 3.13 259 y = 16.607x − 12.958 0.9978 1000–10,000 7 23 850 y = 1.4654x − 1.0688 0.9979 1000–10,000 11 37 Ferulic acid 25.54 321 y = 5.422x + 0.0426 1.0000 125–10,000 33 110 850 y = 0.3652x + 0.0337 0.9986 50–10,000 17 57 3-Hydroxy-benzaldehyde 8.04 255 y = 5.2918x − 2.0066 0.9965 500–5000 249 830 1050 y = 0.3612x − 0.0615 0.9986 500–5000 61 203 4-Hydroxy-benzaldehyde 6.82 284 y = 20.694x − 5.0051 0.9994 500–10,000 36 120 1050 y = 0.499x + 0.5745 0.9757 500–10,000 19 63 iso-Vanillic acid 9.82 259 y = 7.0912x + 0.1118 0.9999 50–10,000 17 57 1050 y = 0.6327x − 0.0102 0.9996 25–10,000 9 30 Salicylic acid 23.43 301 y = 1.1103x + 0.1852 0.9997 250–10,000 102 340 1050 y = 0.2896x + 0.0794 0.9906 50–7500 24 80 Syringic acid 12.34 274 y = 4.5468x + 0.1019 0.9999 50–10,000 22 73 850 y = 0.4868x + 0.0406 0.9991 25–10,000 12 40 Syringaldehyde 16.85 306 y = 4.5868x + 0.0123 0.9999 125–10,000 28 93 900 y = 0.2618x + 0.0716 0.9927 50–10,000 22 73 Vanillic acid 8.28 259 y = 8.0969x + 0.4406 0.9999 25–10,000 15 50 900 y = 0.7441x + 0.0468 0.9992 50–10,000 8 27 Vanillin 11.00 279 y = 6.9652x − 0.3602 0.9999 50–10,000 11 37 900 y = 0.6845x + 0.1581 0.9893 50 − 10,000 15 50 Analytes . DAD . CoulArray .   . RT (min) . λ (nm) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Main channel (mV) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Acetosyringone 26.73 298 y = 3.6395x + 0.1993 0.9994 125–10,000 43 143 850 y = 0.2586x + 0.0612 0.9941 50–10,000 24 80 Acetovanillone 18.55 275 y = 3.3964x + 0.1032 0.9998 50–10,000 29 97 900 y = 0.4124x + 0.054 0.9972 50–10,000 15 50 Benzoic acid 20.50 231 y = 4.4802x − 1.5685 0.999 2500–10,000 477 1590 n.d. n.d. n.d. n.d. n.d. n.d. p-Coumaric acid 16.25 308 y = 9.2165x − 0.0434 1.0000 50–10,000 14 47 900 y = 0.3892x + 0.1408 0.9875 25–10,000 15 50 3.4-Dihydroxy-benzoic acid 3.13 259 y = 16.607x − 12.958 0.9978 1000–10,000 7 23 850 y = 1.4654x − 1.0688 0.9979 1000–10,000 11 37 Ferulic acid 25.54 321 y = 5.422x + 0.0426 1.0000 125–10,000 33 110 850 y = 0.3652x + 0.0337 0.9986 50–10,000 17 57 3-Hydroxy-benzaldehyde 8.04 255 y = 5.2918x − 2.0066 0.9965 500–5000 249 830 1050 y = 0.3612x − 0.0615 0.9986 500–5000 61 203 4-Hydroxy-benzaldehyde 6.82 284 y = 20.694x − 5.0051 0.9994 500–10,000 36 120 1050 y = 0.499x + 0.5745 0.9757 500–10,000 19 63 iso-Vanillic acid 9.82 259 y = 7.0912x + 0.1118 0.9999 50–10,000 17 57 1050 y = 0.6327x − 0.0102 0.9996 25–10,000 9 30 Salicylic acid 23.43 301 y = 1.1103x + 0.1852 0.9997 250–10,000 102 340 1050 y = 0.2896x + 0.0794 0.9906 50–7500 24 80 Syringic acid 12.34 274 y = 4.5468x + 0.1019 0.9999 50–10,000 22 73 850 y = 0.4868x + 0.0406 0.9991 25–10,000 12 40 Syringaldehyde 16.85 306 y = 4.5868x + 0.0123 0.9999 125–10,000 28 93 900 y = 0.2618x + 0.0716 0.9927 50–10,000 22 73 Vanillic acid 8.28 259 y = 8.0969x + 0.4406 0.9999 25–10,000 15 50 900 y = 0.7441x + 0.0468 0.9992 50–10,000 8 27 Vanillin 11.00 279 y = 6.9652x − 0.3602 0.9999 50–10,000 11 37 900 y = 0.6845x + 0.1581 0.9893 50 − 10,000 15 50 n.d., not detectable. aLimit of detection (S/N: 3:1); bLimit of quantitation (S/N: 10:1). Open in new tab Table IV Comparison of DAD and CoulArray Detection Performance (Peak Heights), Applying the Formic Acid Gradient Analytes . DAD . CoulArray .   . RT (min) . λ (nm) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Main channel (mV) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Acetosyringone 26.73 298 y = 3.6395x + 0.1993 0.9994 125–10,000 43 143 850 y = 0.2586x + 0.0612 0.9941 50–10,000 24 80 Acetovanillone 18.55 275 y = 3.3964x + 0.1032 0.9998 50–10,000 29 97 900 y = 0.4124x + 0.054 0.9972 50–10,000 15 50 Benzoic acid 20.50 231 y = 4.4802x − 1.5685 0.999 2500–10,000 477 1590 n.d. n.d. n.d. n.d. n.d. n.d. p-Coumaric acid 16.25 308 y = 9.2165x − 0.0434 1.0000 50–10,000 14 47 900 y = 0.3892x + 0.1408 0.9875 25–10,000 15 50 3.4-Dihydroxy-benzoic acid 3.13 259 y = 16.607x − 12.958 0.9978 1000–10,000 7 23 850 y = 1.4654x − 1.0688 0.9979 1000–10,000 11 37 Ferulic acid 25.54 321 y = 5.422x + 0.0426 1.0000 125–10,000 33 110 850 y = 0.3652x + 0.0337 0.9986 50–10,000 17 57 3-Hydroxy-benzaldehyde 8.04 255 y = 5.2918x − 2.0066 0.9965 500–5000 249 830 1050 y = 0.3612x − 0.0615 0.9986 500–5000 61 203 4-Hydroxy-benzaldehyde 6.82 284 y = 20.694x − 5.0051 0.9994 500–10,000 36 120 1050 y = 0.499x + 0.5745 0.9757 500–10,000 19 63 iso-Vanillic acid 9.82 259 y = 7.0912x + 0.1118 0.9999 50–10,000 17 57 1050 y = 0.6327x − 0.0102 0.9996 25–10,000 9 30 Salicylic acid 23.43 301 y = 1.1103x + 0.1852 0.9997 250–10,000 102 340 1050 y = 0.2896x + 0.0794 0.9906 50–7500 24 80 Syringic acid 12.34 274 y = 4.5468x + 0.1019 0.9999 50–10,000 22 73 850 y = 0.4868x + 0.0406 0.9991 25–10,000 12 40 Syringaldehyde 16.85 306 y = 4.5868x + 0.0123 0.9999 125–10,000 28 93 900 y = 0.2618x + 0.0716 0.9927 50–10,000 22 73 Vanillic acid 8.28 259 y = 8.0969x + 0.4406 0.9999 25–10,000 15 50 900 y = 0.7441x + 0.0468 0.9992 50–10,000 8 27 Vanillin 11.00 279 y = 6.9652x − 0.3602 0.9999 50–10,000 11 37 900 y = 0.6845x + 0.1581 0.9893 50 − 10,000 15 50 Analytes . DAD . CoulArray .   . RT (min) . λ (nm) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Main channel (mV) . Regression equation . R 2 . Linear range (μg/L) . LODa (μg/L) . LOQb (μg/L) . Acetosyringone 26.73 298 y = 3.6395x + 0.1993 0.9994 125–10,000 43 143 850 y = 0.2586x + 0.0612 0.9941 50–10,000 24 80 Acetovanillone 18.55 275 y = 3.3964x + 0.1032 0.9998 50–10,000 29 97 900 y = 0.4124x + 0.054 0.9972 50–10,000 15 50 Benzoic acid 20.50 231 y = 4.4802x − 1.5685 0.999 2500–10,000 477 1590 n.d. n.d. n.d. n.d. n.d. n.d. p-Coumaric acid 16.25 308 y = 9.2165x − 0.0434 1.0000 50–10,000 14 47 900 y = 0.3892x + 0.1408 0.9875 25–10,000 15 50 3.4-Dihydroxy-benzoic acid 3.13 259 y = 16.607x − 12.958 0.9978 1000–10,000 7 23 850 y = 1.4654x − 1.0688 0.9979 1000–10,000 11 37 Ferulic acid 25.54 321 y = 5.422x + 0.0426 1.0000 125–10,000 33 110 850 y = 0.3652x + 0.0337 0.9986 50–10,000 17 57 3-Hydroxy-benzaldehyde 8.04 255 y = 5.2918x − 2.0066 0.9965 500–5000 249 830 1050 y = 0.3612x − 0.0615 0.9986 500–5000 61 203 4-Hydroxy-benzaldehyde 6.82 284 y = 20.694x − 5.0051 0.9994 500–10,000 36 120 1050 y = 0.499x + 0.5745 0.9757 500–10,000 19 63 iso-Vanillic acid 9.82 259 y = 7.0912x + 0.1118 0.9999 50–10,000 17 57 1050 y = 0.6327x − 0.0102 0.9996 25–10,000 9 30 Salicylic acid 23.43 301 y = 1.1103x + 0.1852 0.9997 250–10,000 102 340 1050 y = 0.2896x + 0.0794 0.9906 50–7500 24 80 Syringic acid 12.34 274 y = 4.5468x + 0.1019 0.9999 50–10,000 22 73 850 y = 0.4868x + 0.0406 0.9991 25–10,000 12 40 Syringaldehyde 16.85 306 y = 4.5868x + 0.0123 0.9999 125–10,000 28 93 900 y = 0.2618x + 0.0716 0.9927 50–10,000 22 73 Vanillic acid 8.28 259 y = 8.0969x + 0.4406 0.9999 25–10,000 15 50 900 y = 0.7441x + 0.0468 0.9992 50–10,000 8 27 Vanillin 11.00 279 y = 6.9652x − 0.3602 0.9999 50–10,000 11 37 900 y = 0.6845x + 0.1581 0.9893 50 − 10,000 15 50 n.d., not detectable. aLimit of detection (S/N: 3:1); bLimit of quantitation (S/N: 10:1). Open in new tab The formic acid gradient is less suited for coulometric detection. Compared with the corresponding data compiled in Table III, the coefficients of determination are mostly lower and the detection limits are higher. Nevertheless, CoulArray detection sensitivity remained superior to DAD for the majority of analytes. As Table V reveals, the reproducibility of the analyte quantification by DAD is superior to electrochemical detection. With a mean variation coefficient of 0.76 (all coefficients ≤1.0), the DAD intra-day precision of the phosphate buffer gradient method is slightly better than that calculated for the formic acid gradient. This relation is reversed for the inter-day precision. Inter-day precision, applying the formic acid gradient in combination with DAD detection, is nearly as good (mean value: 1.06) as the corresponding intra-day precision, whereas the relative standard deviations doubled applying the phosphate buffer gradient. Surprisingly, CoulArray intra-day precision was superior for the formic acid eluent. Except for three compounds, variation coefficients were <5.0, whereas this value was exceeded by seven compounds applying the other eluent. CoulArray detection signal provided a low inter-day precision generally. The variation coefficients for the phosphate buffer method were around 10%, mainly with a minimum of approximately 2% (syringaldehyde) and a maximum of 17.9% (3,4-dihydroxybenzoic acid). A higher data scattering is obvious for the formic acid gradient where a maximum RSD of 59.5% was ascertained for acetosyringone. It is noteworthy that the three compounds with the lowest inter-day precision, i.e., acetosyringone, syringic acid and syringaldehyde, respectively, were quantified, selecting the same dominant channel of 850 mV. In contrast to the “ideal” electrochemical behavior, we ascertained considerable variabilities of the peak height and peak size ratio between the main and the subsequent channels at higher potentials. Thus, the substance proportions oxidized at the main channel potential varied from day to day moderately. Table V Intra-Day and Inter-Day Precision of Peak Heights Analyte . Intra-day . Inter-day .   . Formic acid gradient . Phosphate buffer gradient . Formic acid gradient . Phosphate buffer gradient . Detector . DAD . CoulArray . DAD . CoulArray . DAD . CoulArray . DAD . CoulArray . 3.4-Dihydroxybenzoic acid 0.33 1.18 0.51 2.06 0.26 1.25 0.75 17.9 4-Hydroxybenzaldehyde 0.45 4.63 0.81 4.97 0.71 4.62 1.33 10.5 3-Hydroxybenzaldehyde 0.45 3.93 0.58 8.64 0.66 1.34 1.43 6.69 Vanillic acid 0.51 1.82 0.90 1.38 0.78 8.00 1.74 11.5 iso-Vanillic acid 0.70 1.77 0.77 0.47 0.90 13.2 2.10 11.5 Vanillin 0.68 4.60 0.64 6.86 0.87 3.23 1.18 11.3 Syringic acid 0.85 0.78 0.85 1.71 1.05 48.8 1.66 4.06 p-Coumaric acid 0.81 16.2 0.77 8.20 1.21 9.32 1.21 13.6 Syringaldehyde 1.02 6.87 1.00 13.10 1.05 15.6 1.34 1.91 Acetovanillone 1.38 3.83 0.90 8.16 0.99 7.12 1.39 11.9 Benzoic acid 3.00 n.d. 0.74 n.d. 1.52 n.d. 1.72 n.d. Salicylic acid 0.78 8.79 0.76 5.33 0.98 4.04 2.85 8.70 Ferulic acid 0.91 1.10 0.69 5.61 1.87 7.30 1.41 5.43 Acetosyringone 1.09 0.81 0.67 2.09 2.00 59.5 1.45 4.92 Analyte . Intra-day . Inter-day .   . Formic acid gradient . Phosphate buffer gradient . Formic acid gradient . Phosphate buffer gradient . Detector . DAD . CoulArray . DAD . CoulArray . DAD . CoulArray . DAD . CoulArray . 3.4-Dihydroxybenzoic acid 0.33 1.18 0.51 2.06 0.26 1.25 0.75 17.9 4-Hydroxybenzaldehyde 0.45 4.63 0.81 4.97 0.71 4.62 1.33 10.5 3-Hydroxybenzaldehyde 0.45 3.93 0.58 8.64 0.66 1.34 1.43 6.69 Vanillic acid 0.51 1.82 0.90 1.38 0.78 8.00 1.74 11.5 iso-Vanillic acid 0.70 1.77 0.77 0.47 0.90 13.2 2.10 11.5 Vanillin 0.68 4.60 0.64 6.86 0.87 3.23 1.18 11.3 Syringic acid 0.85 0.78 0.85 1.71 1.05 48.8 1.66 4.06 p-Coumaric acid 0.81 16.2 0.77 8.20 1.21 9.32 1.21 13.6 Syringaldehyde 1.02 6.87 1.00 13.10 1.05 15.6 1.34 1.91 Acetovanillone 1.38 3.83 0.90 8.16 0.99 7.12 1.39 11.9 Benzoic acid 3.00 n.d. 0.74 n.d. 1.52 n.d. 1.72 n.d. Salicylic acid 0.78 8.79 0.76 5.33 0.98 4.04 2.85 8.70 Ferulic acid 0.91 1.10 0.69 5.61 1.87 7.30 1.41 5.43 Acetosyringone 1.09 0.81 0.67 2.09 2.00 59.5 1.45 4.92 n.d., not detectable. Open in new tab Table V Intra-Day and Inter-Day Precision of Peak Heights Analyte . Intra-day . Inter-day .   . Formic acid gradient . Phosphate buffer gradient . Formic acid gradient . Phosphate buffer gradient . Detector . DAD . CoulArray . DAD . CoulArray . DAD . CoulArray . DAD . CoulArray . 3.4-Dihydroxybenzoic acid 0.33 1.18 0.51 2.06 0.26 1.25 0.75 17.9 4-Hydroxybenzaldehyde 0.45 4.63 0.81 4.97 0.71 4.62 1.33 10.5 3-Hydroxybenzaldehyde 0.45 3.93 0.58 8.64 0.66 1.34 1.43 6.69 Vanillic acid 0.51 1.82 0.90 1.38 0.78 8.00 1.74 11.5 iso-Vanillic acid 0.70 1.77 0.77 0.47 0.90 13.2 2.10 11.5 Vanillin 0.68 4.60 0.64 6.86 0.87 3.23 1.18 11.3 Syringic acid 0.85 0.78 0.85 1.71 1.05 48.8 1.66 4.06 p-Coumaric acid 0.81 16.2 0.77 8.20 1.21 9.32 1.21 13.6 Syringaldehyde 1.02 6.87 1.00 13.10 1.05 15.6 1.34 1.91 Acetovanillone 1.38 3.83 0.90 8.16 0.99 7.12 1.39 11.9 Benzoic acid 3.00 n.d. 0.74 n.d. 1.52 n.d. 1.72 n.d. Salicylic acid 0.78 8.79 0.76 5.33 0.98 4.04 2.85 8.70 Ferulic acid 0.91 1.10 0.69 5.61 1.87 7.30 1.41 5.43 Acetosyringone 1.09 0.81 0.67 2.09 2.00 59.5 1.45 4.92 Analyte . Intra-day . Inter-day .   . Formic acid gradient . Phosphate buffer gradient . Formic acid gradient . Phosphate buffer gradient . Detector . DAD . CoulArray . DAD . CoulArray . DAD . CoulArray . DAD . CoulArray . 3.4-Dihydroxybenzoic acid 0.33 1.18 0.51 2.06 0.26 1.25 0.75 17.9 4-Hydroxybenzaldehyde 0.45 4.63 0.81 4.97 0.71 4.62 1.33 10.5 3-Hydroxybenzaldehyde 0.45 3.93 0.58 8.64 0.66 1.34 1.43 6.69 Vanillic acid 0.51 1.82 0.90 1.38 0.78 8.00 1.74 11.5 iso-Vanillic acid 0.70 1.77 0.77 0.47 0.90 13.2 2.10 11.5 Vanillin 0.68 4.60 0.64 6.86 0.87 3.23 1.18 11.3 Syringic acid 0.85 0.78 0.85 1.71 1.05 48.8 1.66 4.06 p-Coumaric acid 0.81 16.2 0.77 8.20 1.21 9.32 1.21 13.6 Syringaldehyde 1.02 6.87 1.00 13.10 1.05 15.6 1.34 1.91 Acetovanillone 1.38 3.83 0.90 8.16 0.99 7.12 1.39 11.9 Benzoic acid 3.00 n.d. 0.74 n.d. 1.52 n.d. 1.72 n.d. Salicylic acid 0.78 8.79 0.76 5.33 0.98 4.04 2.85 8.70 Ferulic acid 0.91 1.10 0.69 5.61 1.87 7.30 1.41 5.43 Acetosyringone 1.09 0.81 0.67 2.09 2.00 59.5 1.45 4.92 n.d., not detectable. Open in new tab Spike recoveries were determined by adding known amounts of the analytes to lignin, sawdust and spruce particles before or after the microwave digestion (Table VI). Samples were analyzed by applying the phosphate buffer gradient in combination with both detectors exclusively. The gained recoveries depended strongly on sample type, sample treatment, detector type and on the specific analyte. 3,4-dihydroxybenzoic acid could not be recovered from any sample, if added before digestion. Its recoveries from the digestion solutions were also very low. This might be explained by its low redox potential facilitating a fast oxidative transformation. Comparing the means of the differences between the practical and theoretical recoveries (3,4-dihydroxybenzoic acid excluded), which are calculated for every sample (bottom line of Table VI), it gets obvious that the mean recoveries of the spiked digestion solutions are superior to that of the spiked solid materials, except of the coulometrically quantified lignin samples. Better mean recoveries were achieved with DAD detection with the exception of the spiked sawdust. The detection-depending differences of the mean recoveries were greatest for the spiked lignin hydrolysate. Here, the CoulArray data of several compounds resulted in strong negative as well as positive deviations from the theoretical value. Except ferulic acid, all recoveries from spruce particles, spiked before digestion, were below 80%. Another interesting aspect is the substance-specific variabilities of the recoveries and their dependence on sample pretreatment and detection mode. The DAD recovery of syringic acid is relatively constant and independent from sample type and spiking conditions. If the evaluation is made for the spiking variants separately, relatively uniform recoveries were obtained with both detectors for vanillic acid, iso-vanillic acid, syringaldehyde, ferulic acid and acetovanillone. Mean substance-specific differences between the detector signals (last column, Table VI) were <10% for acetosyringone, salicylic acid, iso-vanillic acid and syringic acid. Highest differences of the detector-specific recoveries were established for vanillin due to very low coulometric recoveries in both lignin samples. Table VI Spike Recoveries from Hydrolysates of Lignin and Lignin-Containing Biomass Sample . Spike addition before sample digestion . Spike addition to the digestion solution .   .   . Lignin . Sawdust . Spruce particles . Lignin . Sawdust . Spruce particles . Mean difference (%) between detectors . Detector . DAD . CA . DAD . CA . DAD . CA . DAD . CA . DAD . CA . DAD . CA .   .   . Recoverya (%) . Recoveryb (%) .   . Analyte 3.4-Dihydroxybenzoic acid 0.12 0.0 0.0 0.0 0.0 0.0 15.7 17.9 45.9 51.6 53.6 37.4 n.q. 4-Hydroxybenzaldehyde 91.9 97.2 93.9 100.3 75.4 51.9 96.3 128.6 95.6 98.2 91.5 70.4 15.2 Vanillic acid 65.4 70.4 65.4 75.0 67.5 65.9 82.5 162.5 82.5 78.2 85.7 77.2 18.2 Salicylic acid 82.8 83.4 89.1 95.9 73.1 66.9 97.7 89.4 95.8 103.8 90.3 87.4 5.5 3-Hydroxybenzaldehyde 70.7 81.2 53.2 58.7 53.2 50.5 98.7 47.2 98.1 107.3 58.0 49.5 14.7 iso-Vanillic acid 89.7 95.3 89.1 84.2 75.7 75.3 91.0 77.0 91.3 87.5 90.9 83.3 6.1 Syringic acid 85.6 88.1 83.2 84.9 74.9 77.6 85.0 45.5 85.7 91.0 89.4 88.2 8.8 Vanillin 76.0c 41.7c 95.4c 100.1c 59.3d 55.9d 94.0e 25.6e 97.2e 106.4e 113.5 111.0 20.4 p-Coumaric acid 82.0 70.4 76.7 87.4 72.6 74.4 88.5 70.6 88.5 102.4 89.2 95.6 10.4 Benzoic acid 83.9 n.d. 86.6 n.d. 76.7 n.d. 82.3 n.d. 85.2 n.d. 90.5 n.d. n.q. Syringaldehyde 86.5 67.7 92.9 79.6 72.8 76.8 91.9e 71.3e 91.9e 75.2e 105.6 105.3 12.3 Acetovanillone 72.0 34.4 92.1 99.5 62.5 66.1 93.1e 91.7e 93.7e 81.6e 93.0 85.1 11.7 Ferulic acid 75.9 83.7 71.5 75.3 68.8 81.6 89.9 77.6 89.7 97.3 89.9 93.2 12.9 Acetosyringone 82.0 80.8 85.7 81.6 76.3 73.0 93.0 83.2 92.3 89.8 90.6 86.7 4.1 Mean deviation from 100% (%) 19.7 24.7 17.3 14.9 30.1 32.8 8.9 34.3 8.4 10.1 12.3 16.7 Sample . Spike addition before sample digestion . Spike addition to the digestion solution .   .   . Lignin . Sawdust . Spruce particles . Lignin . Sawdust . Spruce particles . Mean difference (%) between detectors . Detector . DAD . CA . DAD . CA . DAD . CA . DAD . CA . DAD . CA . DAD . CA .   .   . Recoverya (%) . Recoveryb (%) .   . Analyte 3.4-Dihydroxybenzoic acid 0.12 0.0 0.0 0.0 0.0 0.0 15.7 17.9 45.9 51.6 53.6 37.4 n.q. 4-Hydroxybenzaldehyde 91.9 97.2 93.9 100.3 75.4 51.9 96.3 128.6 95.6 98.2 91.5 70.4 15.2 Vanillic acid 65.4 70.4 65.4 75.0 67.5 65.9 82.5 162.5 82.5 78.2 85.7 77.2 18.2 Salicylic acid 82.8 83.4 89.1 95.9 73.1 66.9 97.7 89.4 95.8 103.8 90.3 87.4 5.5 3-Hydroxybenzaldehyde 70.7 81.2 53.2 58.7 53.2 50.5 98.7 47.2 98.1 107.3 58.0 49.5 14.7 iso-Vanillic acid 89.7 95.3 89.1 84.2 75.7 75.3 91.0 77.0 91.3 87.5 90.9 83.3 6.1 Syringic acid 85.6 88.1 83.2 84.9 74.9 77.6 85.0 45.5 85.7 91.0 89.4 88.2 8.8 Vanillin 76.0c 41.7c 95.4c 100.1c 59.3d 55.9d 94.0e 25.6e 97.2e 106.4e 113.5 111.0 20.4 p-Coumaric acid 82.0 70.4 76.7 87.4 72.6 74.4 88.5 70.6 88.5 102.4 89.2 95.6 10.4 Benzoic acid 83.9 n.d. 86.6 n.d. 76.7 n.d. 82.3 n.d. 85.2 n.d. 90.5 n.d. n.q. Syringaldehyde 86.5 67.7 92.9 79.6 72.8 76.8 91.9e 71.3e 91.9e 75.2e 105.6 105.3 12.3 Acetovanillone 72.0 34.4 92.1 99.5 62.5 66.1 93.1e 91.7e 93.7e 81.6e 93.0 85.1 11.7 Ferulic acid 75.9 83.7 71.5 75.3 68.8 81.6 89.9 77.6 89.7 97.3 89.9 93.2 12.9 Acetosyringone 82.0 80.8 85.7 81.6 76.3 73.0 93.0 83.2 92.3 89.8 90.6 86.7 4.1 Mean deviation from 100% (%) 19.7 24.7 17.3 14.9 30.1 32.8 8.9 34.3 8.4 10.1 12.3 16.7 CA, CoulArray; n.d., not detectable; n.q., not quantifiable. aSpike amount 1 mg, if not stated otherwise; beffective spike concentration: 1.0 mg/L, if not stated otherwise; cspike amount 5 mg; dspike amount: 4 mg; eeffective spike concentration: 5.0 mg/L. Open in new tab Table VI Spike Recoveries from Hydrolysates of Lignin and Lignin-Containing Biomass Sample . Spike addition before sample digestion . Spike addition to the digestion solution .   .   . Lignin . Sawdust . Spruce particles . Lignin . Sawdust . Spruce particles . Mean difference (%) between detectors . Detector . DAD . CA . DAD . CA . DAD . CA . DAD . CA . DAD . CA . DAD . CA .   .   . Recoverya (%) . Recoveryb (%) .   . Analyte 3.4-Dihydroxybenzoic acid 0.12 0.0 0.0 0.0 0.0 0.0 15.7 17.9 45.9 51.6 53.6 37.4 n.q. 4-Hydroxybenzaldehyde 91.9 97.2 93.9 100.3 75.4 51.9 96.3 128.6 95.6 98.2 91.5 70.4 15.2 Vanillic acid 65.4 70.4 65.4 75.0 67.5 65.9 82.5 162.5 82.5 78.2 85.7 77.2 18.2 Salicylic acid 82.8 83.4 89.1 95.9 73.1 66.9 97.7 89.4 95.8 103.8 90.3 87.4 5.5 3-Hydroxybenzaldehyde 70.7 81.2 53.2 58.7 53.2 50.5 98.7 47.2 98.1 107.3 58.0 49.5 14.7 iso-Vanillic acid 89.7 95.3 89.1 84.2 75.7 75.3 91.0 77.0 91.3 87.5 90.9 83.3 6.1 Syringic acid 85.6 88.1 83.2 84.9 74.9 77.6 85.0 45.5 85.7 91.0 89.4 88.2 8.8 Vanillin 76.0c 41.7c 95.4c 100.1c 59.3d 55.9d 94.0e 25.6e 97.2e 106.4e 113.5 111.0 20.4 p-Coumaric acid 82.0 70.4 76.7 87.4 72.6 74.4 88.5 70.6 88.5 102.4 89.2 95.6 10.4 Benzoic acid 83.9 n.d. 86.6 n.d. 76.7 n.d. 82.3 n.d. 85.2 n.d. 90.5 n.d. n.q. Syringaldehyde 86.5 67.7 92.9 79.6 72.8 76.8 91.9e 71.3e 91.9e 75.2e 105.6 105.3 12.3 Acetovanillone 72.0 34.4 92.1 99.5 62.5 66.1 93.1e 91.7e 93.7e 81.6e 93.0 85.1 11.7 Ferulic acid 75.9 83.7 71.5 75.3 68.8 81.6 89.9 77.6 89.7 97.3 89.9 93.2 12.9 Acetosyringone 82.0 80.8 85.7 81.6 76.3 73.0 93.0 83.2 92.3 89.8 90.6 86.7 4.1 Mean deviation from 100% (%) 19.7 24.7 17.3 14.9 30.1 32.8 8.9 34.3 8.4 10.1 12.3 16.7 Sample . Spike addition before sample digestion . Spike addition to the digestion solution .   .   . Lignin . Sawdust . Spruce particles . Lignin . Sawdust . Spruce particles . Mean difference (%) between detectors . Detector . DAD . CA . DAD . CA . DAD . CA . DAD . CA . DAD . CA . DAD . CA .   .   . Recoverya (%) . Recoveryb (%) .   . Analyte 3.4-Dihydroxybenzoic acid 0.12 0.0 0.0 0.0 0.0 0.0 15.7 17.9 45.9 51.6 53.6 37.4 n.q. 4-Hydroxybenzaldehyde 91.9 97.2 93.9 100.3 75.4 51.9 96.3 128.6 95.6 98.2 91.5 70.4 15.2 Vanillic acid 65.4 70.4 65.4 75.0 67.5 65.9 82.5 162.5 82.5 78.2 85.7 77.2 18.2 Salicylic acid 82.8 83.4 89.1 95.9 73.1 66.9 97.7 89.4 95.8 103.8 90.3 87.4 5.5 3-Hydroxybenzaldehyde 70.7 81.2 53.2 58.7 53.2 50.5 98.7 47.2 98.1 107.3 58.0 49.5 14.7 iso-Vanillic acid 89.7 95.3 89.1 84.2 75.7 75.3 91.0 77.0 91.3 87.5 90.9 83.3 6.1 Syringic acid 85.6 88.1 83.2 84.9 74.9 77.6 85.0 45.5 85.7 91.0 89.4 88.2 8.8 Vanillin 76.0c 41.7c 95.4c 100.1c 59.3d 55.9d 94.0e 25.6e 97.2e 106.4e 113.5 111.0 20.4 p-Coumaric acid 82.0 70.4 76.7 87.4 72.6 74.4 88.5 70.6 88.5 102.4 89.2 95.6 10.4 Benzoic acid 83.9 n.d. 86.6 n.d. 76.7 n.d. 82.3 n.d. 85.2 n.d. 90.5 n.d. n.q. Syringaldehyde 86.5 67.7 92.9 79.6 72.8 76.8 91.9e 71.3e 91.9e 75.2e 105.6 105.3 12.3 Acetovanillone 72.0 34.4 92.1 99.5 62.5 66.1 93.1e 91.7e 93.7e 81.6e 93.0 85.1 11.7 Ferulic acid 75.9 83.7 71.5 75.3 68.8 81.6 89.9 77.6 89.7 97.3 89.9 93.2 12.9 Acetosyringone 82.0 80.8 85.7 81.6 76.3 73.0 93.0 83.2 92.3 89.8 90.6 86.7 4.1 Mean deviation from 100% (%) 19.7 24.7 17.3 14.9 30.1 32.8 8.9 34.3 8.4 10.1 12.3 16.7 CA, CoulArray; n.d., not detectable; n.q., not quantifiable. aSpike amount 1 mg, if not stated otherwise; beffective spike concentration: 1.0 mg/L, if not stated otherwise; cspike amount 5 mg; dspike amount: 4 mg; eeffective spike concentration: 5.0 mg/L. Open in new tab Application to samples Figure 4 presents an example of a DAD chromatogram of a diluted sawdust digest, applying the phosphate buffer gradient. Main components are vanillin and syringaldehyde, followed by acetovanillone and acetosyringone. Vanillic acid was detected in traces. The digestion of a lignin sample (data not shown) resulted in essentially the same vanillin and acetovanillone amounts, but roughly one order of magnitude lower syringaldehyde and acetosyringone contents, and in an elevated vanillic acid yield. Parallel quantitation by coulometric detection demonstrated a good agreement of both data sets. Very small peaks eluting in the retention time windows of 3,4-dihydroxybenzoic acid, 3-hydroxybenzaldehyde and 4-hydroxybenzaldehyde could not be clearly assigned since their DAD spectra and hydrodynamic voltammograms did not sufficiently match those of the reference compounds. Figure 4 Open in new tabDownload slide Chromatogram of a sawdust digest; peaks: 1: 3,4-dihydroxybenzoic acid, 2: 4-hydroxybenzaldehyde, 3: vanillic acid, 4: iso-vanillic acid, 5: vanillin, 6: syringaldehyde, 7: acetovanillone and 8: acetosyringone; phosphate buffer gradient, DAD: 231 nm. Figure 4 Open in new tabDownload slide Chromatogram of a sawdust digest; peaks: 1: 3,4-dihydroxybenzoic acid, 2: 4-hydroxybenzaldehyde, 3: vanillic acid, 4: iso-vanillic acid, 5: vanillin, 6: syringaldehyde, 7: acetovanillone and 8: acetosyringone; phosphate buffer gradient, DAD: 231 nm. Discussion Reflecting the achieved results, one has to take into account that the pursued aim was to develop a versatile analytical method capable to determine relevant monomeric products of lignin hydrolysis. Thus, the composition of the group of analytes, including phenolic acids, aldehydes and ketones, differed from that of otherwise similar methodical challenges. The analytical task defined specific requirements on the chromatographic and detection selectivity also. The three-step phosphate buffer gradient enabled the separation of the 14 analytes within 28 min. Three peak pairs were not completely resolved, but resolution was sufficient for uncompromised quantification. Two out of the three incompletely resolved pairs were composed of a phenolic acid and an aldehyde. In these cases, complete separation would be possible, lowering the eluent pH leading to an enhanced retention of the acidic component, but at the expense of worsening the resolution in other sections of the chromatogram. Neither pH nor temperature shifts improved the separation between syringaldehyde and acetovanillone. Despite of these shortcomings, the application of the simple but highly reproducible phosphate gradient (RSD of the retention time of the majority of the analytes ≤0.5%) in combination with the Kinetex RP column achieved comparable or superior chromatographic performance compared with more complex and less robust methods. For instance, Floridi et al. (22) applied a 17-step composition and flow gradient on a RP C18-column to separate 19 analytes (18 phenolic acids plus tyrosol) within 47 min, but 5 peak pairs remained incompletely resolved. Paralleling the previous work of Montanari et al. (24), the method required the addition of a surfactant to the eluent, making the oxidative substance transformation at the electrode surfaces more susceptible to faults. Lobes et al. (12) separated 13 lignin phenol standards including 9 compounds that were identical with those selected here on a Lichrosphere 100 RP 18 column. Applying a binary phosphate gradient, the separation was accomplished within 50 min. Retention time difference between the 4-hydroxybenzaldehyde and vanillic acid was exactly the same as in the here-described method (0.6 min), but elution of this analyte pair occurred in the interval between 22 and 23 min compared with an elution between 10 and 11 min reported here. Further advantages of the here-presented method are the low baseline noise and high optical transparency of the phosphate buffer gradient, enabling DAD detection limits ≤20 μg/L for the majority of analytes. The separation was slightly improved, replacing the phosphate by the formic acid gradient. Two peak pairs could not be completely resolved, but two of the substance pairs, remaining incompletely resolved with the phosphate buffer gradient, were completely separated. Except the higher retention of salicylic acid, the elution sequence resembles the phosphate gradient. Nevertheless, the differences in the retention times, related to the phosphate eluent, are significant, thereby offering complementary chromatographic information to enhance the certainty of analyte identification. The elaborated binary formic acid gradient provides a straightforward alternative to other formic acid gradients, especially in cases where a combination of DAD and MS detection is of interest (16). Canas et al. (25) applied a six-step formic acid/methanol gradient to separate phenolic acids and aldehydes together with furfural and furfural derivatives in brandy samples. Despite comparably high retention times, syringic acid and vanillin coeluted completely. The dissociation of the carboxylic acid groups is strongly depressed at the low pH value maintained by the formic acid gradient. Thus, analyte retention is not influenced by direct ionic interactions. The comparable high retardation of benzoic acid indicates that retention is not only depending on analyte lipophilicity. The elution sequence is ruled by several structural features with positive and negative effects on retention. As already reported by several authors, the methoxy group attached to the benzene ring enhances retention (22, 26). For instance, ferulic acid with one methoxy group elutes after p-coumaric acid (no methoxy group), and syringaldehyde (two methoxy groups) elutes after vanillin (one methoxy group). Hájek et al. (27) noticed an increased retention of compounds with an acrylic acid moiety. This is reflected by the higher retention time of ferulic acid related to vanillin. The retention of the analytes is negatively correlated with their ability to form hydrogen bonds. Thus, retention decreases according to the sequences acetosyringone > syringaldehyde > syringic acid and acetovanillone > vanillin > vanillic acid. Selectivity and sensitivity of the electrochemical detection depend both on the structural analyte features and on the applied analytical conditions, e.g., composition and pH value of the mobile phase. Three dominant channels were identified to offer the best compromise between selectivity and sensitivity with both gradients. The main channels best suited for the combination with the formic acid gradient were higher than those used with the phosphate gradient, presumably caused by the lower pH value of the former. Kahoun et al. (28) investigated the pH influence on the coulometric detection of vanillic acid at various oxidation potentials within the pH range of 2.80–3.35. The released current amounts increased with decreasing pH at 800 mV, but the pH effect reversed at 900 mV. In agreement with other studies, the catechol 3,4-dihydroxybenzoic acid was quantified at a comparably low potential and salicylic acid at a high potential (22, 27, 28). Conclusion Applying two different HPLC gradients and two detection modes, various monomeric phenolic compounds, representative for lignin hydrolysates, were sensitively and reliably determined. The separation performance of the gradients, both used together with a Kinetex RP C18 separation column, was comparable, but the phosphate buffer gradient offered a somewhat higher sensitivity in combination with both detectors. With its higher volatility, the formic acid gradient might be the first choice for MS applications. The combined operation of the DAD and the CoulArray detector resulted in an enhanced certainty of analyte identification. The coulometric detector achieved a moderately higher sensitivity, especially in combination with the phosphate gradient. When used in combination with the formic acid gradient, the linearity of the calibration functions was compromised for a few analytes. The practical appropriateness of the elaborated method could be confirmed by analyzing the various hydrolysates of lignin-containing samples. References 1. Sjöstrom , E. ; Wood chemistry: fundamentals and applications . Academic Press , Cambridge, MA , ( 1993 ). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 2. Gosselink , R.J.A. , de Jong , E., Guran , B., Abacherli , A.; Coordination network for lignin-standardisation, production and applications adapted to market requirements (EU ROLIGNIN) ; Industrial Crops Production , ( 2004 ); 20 : 121 – 129 . Google Scholar Crossref Search ADS WorldCat 3. Landner , L. , Grahn , O., Härdig , J., Lehtinen , K.J., Monfelt , C., Tana , J.; A field study of environmental impacts at a bleached kraft pulp mill site on the Baltic Sea coast ; Ecotoxicology and Environmental Safety , ( 1994 ); 27 : 128 – 157 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Anonymous ; Toxicity and environmental chemistry of wastewater from a kraft and paper mill: fish toxicity studies . AECV (Series) 87-R4 . Alberta Environmental Centre , Alta , Vegreville , ( 1987 ). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 5. Chen , H. ; Lignocellulose biorefinery engineering. Principles and applications . Woodhead Publishing Limited (Elsevier) , Sawston, Cambridge, UK , ( 2015 ). Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 6. Katahira , R. , Mittal , A., McKinney , K., Chen , X., Tucker , M.P., Johnson , D.K. et al. ; Base-catalyzed depolymerization of biorefinery lignins ; ACS Sustainable Chemistry & Engineering , ( 2016 ); 4 : 1474 – 1486 . Google Scholar Crossref Search ADS WorldCat 7. Srinivas , K. , de Carvalho Oliveira , F., Teller , P.J., Gonçalves , A.R., Helms , G.L., Kaer Ahring , B.; Oxidative degradation of biorefinery lignin obtained after pretreatment of forest residues of Douglas Fir ; Bioresource Technology , ( 2016 ); 221 : 394 – 404 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Chen , C. , Jin , D., Ouyang , X., Zhao , L., Qiu , X., Wang , F.; Effect of structural characteristics on the depolymerization of lignin into phenolic monomers ; Fuel , ( 2018 ); 223 : 366 – 372 . Google Scholar Crossref Search ADS WorldCat 9. Abdelaziz , O.Y. , Li , K., Tunå , P., Hulteberg , C.P.; Continuous catalytic depolymerisation and conversion of industrial kraft lignin into low-molecular-weight aromatics ; Biomass Conversion and Biorefinery , ( 2018 ); 8 : 455 – 470 . Google Scholar Crossref Search ADS WorldCat 10. Khoddami , A. , Wilkes , M.A., Roberts , T.H.; Techniques for analysis of plant phenolic compounds ; Molecules , ( 2013 ); 18 : 2328 – 2375 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Capriotti , A.L. , Cavaliere , C., Foglia , P., Piovesana , S., Ventura , S.; Chromatographic methods coupled to mass spectrometry detection for the determination of phenolic acids in plants and fruits ; Journal of Liquid Chromatography & Related Technologies , ( 2015 ); 38 : 353 – 370 . Google Scholar Crossref Search ADS WorldCat 12. Lobes , J.M. , Fitznar , H.P., Kattner , G.; High-performance liquid chromatography of lignin-derived phenols in environmental samples with diode array detection ; Analytical Chemistry , ( 1999 ); 71 : 3008 – 3012 . Google Scholar Crossref Search ADS PubMed WorldCat 13. Casella , I.C. , Ritorti , M.; Chromatographic separations and liquid phase extraction/concentration of some polyphenolic compounds: a comparison between reverse phase and cation-exchange chromatography ; Analytical Methods , ( 2011 ); 3 : 575 – 581 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Mansour , F.R. , Kirkpatrick , C.I., Danielson , N.D.; Ion exclusion chromatography of aromatic acids ; Journal of Chromatographic Science , ( 2013 ); 51 : 655 – 665 . Google Scholar Crossref Search ADS PubMed WorldCat 15. Wang , R. , Yang , H.-J., Yang , X., Cao , B.-H.; Four phenolic acids determined by an improved HPLC method with a programmed ultraviolet wavelength detection and their relationships with lignin content in 13 agricultural residue feeds ; Journal of the Science of Food and Agriculture , ( 2013 ); 93 : 53 – 60 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Garcia , R. , Soares , B., Dias , C.B., Freitas , A.M.C., Cabrita , M.J.; Phenolic and furanic compounds of Portuguese chestnut and French, American and Portuguese oak wood chips ; European Food Research and Technology , ( 2012 ); 235 : 457 – 467 . Google Scholar Crossref Search ADS WorldCat 17. Zheng , M. , Gu , S., Chen , J., Luo , Y., Li , W., Ni , J. et al. ; Development and validation of a sensitive UPLC–MS/MS instrumentation and alkaline nitrobenzene oxidation method for the determination of lignin monomers in wheat straw ; Journal of Chromatography B , ( 2017 ); 1055 : 178 – 184 . Google Scholar Crossref Search ADS WorldCat 18. Vanbeneden , N. , Delvaux , F., Delvaux , F.R.; Determination of hydroxycinnamic acids and volatile phenols in wort and beer by isocratic high-performance liquid chromatography using electrochemical detection ; Journal of Chromatography A , ( 2006 ); 1136 : 237 – 242 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Cantalapiedra , A. , Gismera , M.J., Sevilla , M.T., Procopio , J.R.; Sensitive and selective determination of phenolic compounds from aromatic plants using an electrochemical detection coupled with HPLC method ; Phytochemical Analysis , ( 2014 ); 25 : 247 – 254 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Svendsen , C.N. ; Multi-electrode array detectors in high-performance liquid chromatography: A new dimension in electrochemical analysis ; Analyst , ( 1993 ); 188 : 123 – 129 . Google Scholar Crossref Search ADS WorldCat 21. Peñalvo , J.L. , Nurmi , T.; Application of coulometric electrode array detection to the analysis of isoflavonoids and lignans ; Journal of Pharmaceutical and Biomedical Analysis , ( 2006 ); 41 : 1497 – 1507 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Floridi , S. , Montanari , L., Marconi , O., Fantozzi , P.; Determination of free phenolic acids in wort and beer by coulometric array detection ; Journal of Agricultural and Food Chemistry , ( 2003 ); 51 : 1548 – 1554 . Google Scholar Crossref Search ADS PubMed WorldCat 23. Aaby , K. , Hvattum , E., Skrede , G.; Analysis of flavonoids and other phenolic compounds using high-performance liquid chromatography with coulometric array detection: Relationship to antioxidant activity ; Journal of Agricultural and Food Chemistry , ( 2004 ); 52 : 4595 – 4603 . Google Scholar Crossref Search ADS PubMed WorldCat 24. Montanari , L. , Perretti , G., Patella , F., Giudi , A., Fantozzi , P.; Organic and phenolic acids in beer ; Journal of Agricultural and Food Chemistry , ( 1999 ); 32 : 1 – 5 . Google Scholar OpenURL Placeholder Text WorldCat 25. Canas , S. , Belchior , A.P., Spranger , M.I., Bruno-de-Sousa , R.; High-performance liquid chromatography method for analysis of phenolic acids, phenolic aldehydes, and furanic derivatives in brandies: development and validation ; Journal of Separation Science , ( 2003 ); 26 : 496 – 502 . Google Scholar Crossref Search ADS WorldCat 26. Guo , C. , Cao , G., Sofic , E., Prior , R.L.; High-performance liquid chromatography coupled with coulometric array detection of electroactive components in fruits and vegetables: relationship to oxygen radical absorbance capacity ; Journal of Agricultural and Food Chemistry , ( 1997 ); 45 : 1787 – 1796 . Google Scholar Crossref Search ADS WorldCat 27. Hájek , T. , Skeriková , V., Cesla , P., Vynuchalová , K., Jandera , P.; Multidimensional LC×LC analysis of phenolic and flavone antioxidants with UV-electrochemical coulometric and MS detection ; Journal of Separation Science , ( 2008 ); 31 : 3309 – 3328 . Google Scholar Crossref Search ADS PubMed WorldCat 28. Kahoun , D. , Rezková , S., Veskrnová , K., Královsky , J., Holcapek , M.; Determination of phenolic compounds and hydroxymethylfurfural in meads using high performance liquid chromatography with coulometric-array and UV detection ; Journal of Chromatography A , ( 2008 ); 1202 : 19 – 33 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - RP-HPLC Analysis of Phenolic Lignin Monomers with DAD and Coulometric Array Detection
JF - Journal of Chromatographic Science
DO - 10.1093/chromsci/bmaa105
DA - 2020-12-19
UR - https://www.deepdyve.com/lp/oxford-university-press/rp-hplc-analysis-of-phenolic-lignin-monomers-with-dad-and-coulometric-UYZE8VQbEx
SP - 1
EP - 1
VL - Advance Article
IS - 
DP - DeepDyve
ER -