Crucial correction of the Faix method for the determination of carbonyls in pyrolysis and hydrothermal liquefaction bio-oils with high nitrogen content

Scientific articles |  | REFINE LabInstrumentation
Pyrolysis, FTIR Spectroscopy, Titration
Industries
Energy & Chemicals
Manufacturer

Importance of the topic

Bio-oils produced by fast pyrolysis and hydrothermal liquefaction (HTL) are complex, oxygenated liquids considered as intermediates for renewable fuels. Reactive carbonyl groups (aldehydes and ketones) can constitute a large fraction of these oils and govern storage stability (polymerization and aging) as well as the required operating strategy for subsequent hydrotreatment. Accurate quantification of total carbonyl content is therefore essential for predicting shelf life and for designing hydroprocessing catalysts and conditions. Existing routine quantification using the Faix oximation/titration approach (standardized as ASTM E3146) was developed for woody pyrolysis oils and can fail when applied to bio-oils with substantial nitrogen-containing bases, producing systematic underestimation of carbonyls. This paper identifies the interfering nitrogen chemistries, quantifies their impact, and proposes practical correction procedures to restore reliable carbonyl measurements.

Objectives and study overview

  • Identify nitrogen-containing compounds in bio-oils that interfere with the Faix titrimetric carbonyl assay.
  • Quantify the degree of interference across a diverse set of bio-oils (pyrolysis and HTL) and hydrotreated products with wide-ranging nitrogen contents.
  • Develop and validate simple correction approaches to compensate for basic nitrogen that consumes HCl liberated during oximation, which otherwise leads to underreported carbonyl values.
  • Provide a practical recommendation and an easy-to-use empirical correction for routine laboratories.

Methodology and experimental design

  • Samples: 13 crude bio-oils (from pyrolysis and HTL of 10 feedstocks) and 5 laboratory hydrotreated HTL oils derived from sewage sludge; nitrogen content varied from ~0.75 wt% up to ~6.6 wt% (and higher in some literature cases).
  • Model screening: 65 nitrogen-containing compounds representative of species reported in bio-oils were tested for reactivity in titrimetric assays; 18 compounds were identified as interferents.
  • Spiking experiments: an HTL straw bio-oil (3.80 mmol C=O/g, 0.75 wt% N) was spiked with a 13-component nitrogen mixture at levels producing ~5, 10 and 14 wt% N to demonstrate the progressive underestimation of carbonyls with increasing basic nitrogen.
  • Carbonyl assay: measurements followed ASTM E3146 (oximation with hydroxylamine HCl, triethanolamine as base, DMSO cosolvent, back-titration of excess triethanolamine with standardized HCl).
  • Correction strategies tested (three approaches):
    1. Direct titration of the sample in 80% ethanol with standardized HCl to quantify reactive basic species (simple titration correction).
    2. ASTM E3146 blank-based correction (Blank B vs Blank C variant): run a blank sample without hydroxylamine to capture basic HCl-consuming species after the same 2 h, 80 °C treatment; calculate correction from equivalence point differences.
    3. "Solution B + sample" method: titrate the sample after adding a known aliquot of triethanolamine solution B (the reagent used in the oximation) and calculate how much of the added triethanolamine was consumed by sample basicity; this isolates basic interference more sensitively for low levels.
  • Instrumentation and titrations: potentiometric non-aqueous titration (Metrohm Titrando 888) with a combined non-aqueous pH electrode; basic nitrogen quantified by perchloric acid titration in glacial acetic acid (UOP 269 agreement); HCl titrant standardized and used for back-titrations.
  • Statistics: triplicate (or greater) analyses, combined standard deviation used for corrected values, and Welch’s t-test (α=0.05) to assess significance of correction effects.

Used instrumentation

  • Potentiometric titrator: Metrohm Titrando 888 (for non-aqueous potentiometric titrations).
  • Combined electrode: LL-Solvotrode easy clean (suitable for non-aqueous media).
  • Reagents and solvents: hydroxylamine hydrochloride (oximation reagent), triethanolamine solution B, dimethyl sulfoxide (DMSO) cosolvent, 80% ethanol (dilution solvent), standardized 0.1 M HCl for back titration, 0.02 M perchloric acid in glacial acetic acid for basic N determination.
  • Reaction conditions: oximation at 80 °C, 2 h stirring (ASTM E3146 protocol).

Main results and discussion

  • Interfering nitrogen species: of 65 tested nitrogen compounds, 18 consumed titrant under assay conditions and therefore interfered with carbonyl determination. Major interfering classes were:
    • Aliphatic primary amines (e.g., hexylamine, cyclohexylamine).
    • Cyclic aliphatic secondary amines formed during hydrogenation of heterocycles (e.g., piperidine, decahydroquinoline).
    • Highly alkyl-substituted N-heterocycles where alkylation raises electron density and basicity (e.g., 2,4,6-trimethylpyridine).
  • Impact magnitude: interference increases with total nitrogen content and the fraction of basic/reactive N. For some HTL and hydrotreated samples, measured carbonyls were underestimated by up to ~18% (absolute correction ~0.53 mmol/g in the worst studied case) and in extreme cases yielded zero or even negative apparent carbonyl values when true carbonyl concentration was low (<0.5 mmol/g) and reactive N fraction was large.
  • Hydrotreatment exacerbation: hydrotreatment saturates heterocyclic rings and can convert non-basic heterocycles into basic secondary/primary amines and ammonia, increasing assay interference in hydrotreated oils. Consequently, hydrotreated HTL oils with moderate N can show substantial underreporting of carbonyls if correction is omitted.
  • Comparison of correction methods: all three correction approaches produced broadly consistent corrected carbonyl values. Practical differences:
    • Direct titration: simplest and fastest, but limited sensitivity for samples with low basicity unless larger sample mass is used.
    • ASTM blank-based correction: most laborious, susceptible to higher variability due to multiple pipetting and transfer steps.
    • "Solution B + sample": best compromise—fast, repeatable, sensitive for small amounts of reactive bases, and recommended for routine adoption.
  • Empirical nitrogen-based correction: the authors observed an approximate linear relationship between correction magnitude (mmol C=O/g) and bulk nitrogen content (wt%). The suggested empirical correction is:
    [corrected carbonyl (mmol/g)] = [measured carbonyl (mmol/g)] + 0.08 × [N (wt%)]
    This approach captures the average behavior across the diverse sample set (high R2 in the study), and is particularly useful when nitrogen content exceeds ~2 wt% and a rapid first-order correction is desirable. Limitations: composition-specific deviations occur when the fraction or speciation of interfering nitrogen deviates substantially from the study average (outliers were observed for atypical samples, e.g., bio-oil containing significant non-biomass cofeed like polyurethane).
  • Statistical validation: Welch’s t-test showed statistically significant differences between uncorrected and corrected carbonyl values for many samples, especially those with N > 2 wt%. For samples with low N (<2 wt%) corrections were often within measurement repeatability.

Benefits and practical applications of the corrected method

  • Improved accuracy in carbonyl quantification enables more reliable accelerated-aging tests and shelf-life predictions for bio-oils.
  • Better-informed hydrotreatment design: knowing true reactive carbonyl loads allows correct staging and temperature selection to avoid polymerization and catalyst fouling.
  • Routine laboratory feasibility: the recommended "solution B + sample" correction requires no new instrumentation and adds a brief titration step, making it straightforward to adopt in QC laboratories that already follow ASTM E3146.
  • Mitigates risk of false-negative carbonyl results that could otherwise lead to inappropriate processing decisions and downstream operational problems.

Future trends and possibilities of use

  • Standardization: the authors recommend incorporating a correction for basic nitrogen into ASTM E3146 to make the standard robust across diverse feedstocks and upgraded products.
  • Speciation-driven corrections: coupling bulk titration correction with complementary nitrogen speciation (e.g., GC-MS of derivatized fractions, NMR, or high-resolution MS) could refine corrections for bio-oils with atypical nitrogen distributions.
  • Automated workflows: automated titration sequences that perform the solution B + sample correction concurrently with the oximation could streamline routine QC.
  • Alternative analytical approaches: development or broader adoption of orthogonal carbonyl assays (e.g., 19F-NMR after derivatization, selective chromatographic quantification) could provide confirmatory data and reduce reliance on empirical corrections.
  • Electrode/solvent optimization: further work on electrode types and non-aqueous solvents that reduce emulsion artifacts for hydrotreated oils may improve repeatability.

Conclusion

  • The Faix/Astm E3146 titrimetric carbonyl assay underestimates carbonyl content in bio-oils that contain significant amounts of basic nitrogen compounds because these species consume the HCl released during oximation.
  • Screening of 65 model nitrogen species identified 18 interferents, mainly primary aliphatic amines, hydrogenation-derived secondary amines, and strongly alkylated heterocycles.
  • Three correction approaches were evaluated; the titration of the sample in the presence of triethanolamine solution B was identified as the most practical and reproducible for routine use.
  • For rapid correction, an empirical formula ([corrected C=O] = [measured C=O] + 0.08 × [N wt%]) provides a useful first-order adjustment, particularly when N > 2 wt%.
  • Adoption of a correction step will prevent gross underestimation (including zero or negative apparent carbonyls) and improve decision-making for storage and hydrotreatment of bio-oils, especially those derived from protein-rich or sludge feedstocks.

Reference

  1. C. Lindfors et al., Energy & Fuels, 37 (2023) 799.
  2. A. Paiva Pinheiro Pires et al., Energy & Fuels (2023).
  3. S. Badoga et al., Fuel, 324 (2022) 124608.
  4. M. Staš et al., TrAC Trends in Analytical Chemistry (2020) 115857.
  5. A. Oasmaa et al., Energy & Fuels, 25 (2011) 3307.
  6. R.N. Hilten and K.C. Das, Fuel, 89 (2010) 2741.
  7. S. Black and J.R. Ferrell, RSC Advances, 10 (2020) 10046.
  8. Y. Han et al., Fuel Processing Technology, 195 (2019) 106140.
  9. B. Shumeiko et al., Fuel Processing Technology, 222 (2021) 106977.
  10. C. Lievens et al., Fuel, 90 (2011) 3417.
  11. F. Huang et al., RSC Advances, 4 (2014) 17743.
  12. O. Faix et al., Holzforschung, 52 (1998) 268.
  13. S. Black and J.R. Ferrell III, Energy & Fuels, 30 (2016) 1071.
  14. ASTM E3146, ASTM International, West Conshohocken, PA (2018).
  15. C. Lindfors et al., Energy & Fuels, 37 (2023) 4460.
  16. R. Ghadge et al., Energy Conversion and Management: X, 14 (2022) 100223.
  17. D. Xu et al., Renewable and Sustainable Energy Reviews, 97 (2018) 103.
  18. S. Nagappan et al., Fuel, 285 (2021) 119053.
  19. M. Gholizadeh et al., Renewable and Sustainable Energy Reviews, 114 (2019) 109313.
  20. L. Leng et al., Chemical Engineering Journal, 401 (2020) 126030.
  21. L. Leng et al., Fuel, 335 (2023) 126995.
  22. W. Chen et al., Environmental Science & Technology, 51 (2017) 6570.
  23. A. Eschenbacher et al., Fuel, 302 (2021) 121198.
  24. Y. Huang et al., Fuel, 183 (2016) 9.
  25. H.E. Toraman et al., Journal of Chromatography A, 1460 (2016) 135.
  26. J. Zimmermann et al., Energy & Fuels, 36 (2022) 14292.
  27. X. Huang et al., Journal of Analytical and Applied Pyrolysis, 110 (2014) 353.
  28. A. Jaramillo-Arango et al., Journal of Analytical and Applied Pyrolysis, 121 (2016) 287.
  29. W. Costanzo et al., Algal Research, 13 (2016) 53.
  30. P. Biller and A.B. Ross, Bioresource Technology, 102 (2011) 215.
  31. M.E. Sánchez et al., Biomass and Bioenergy, 33 (2009) 933.
  32. J. Souza dos Passos et al., Journal of Environmental Chemical Engineering, 10 (2022) 108181.
  33. Q. Niu et al., Journal of Analytical and Applied Pyrolysis, 155 (2021) 105068.
  34. T. Schulzke et al., Waste and Biomass Valorization, 11 (2020) 579.
  35. K. Anastasakis et al., Energies (2018).
  36. J. Souza dos Passos et al., ACS Sustainable Chemistry & Engineering, 8 (2020) 19051.
  37. J. Souza dos Passos et al., Chemical Engineering Journal, 452 (2023) 139636.
  38. C. Hognon et al., Biomass and Bioenergy, 73 (2015) 23.

Content was automatically generated from an orignal PDF document using AI and may contain inaccuracies.

Downloadable PDF for viewing
 

Similar PDF

Upgrading of hydrothermal liquefaction biocrudes from mono- and co-liquefaction of cow manure and wheat straw through hydrotreating and distillation
Bio-based refinery intermediate production via hydrodeoxygenation of fast pyrolysis bio-oil
Efficient One-Stage Bio-Oil Upgrading over Sulfided Catalysts
Quantitative Study of Straw Bio-oil Hydrodeoxygenation over a Sulfided NiMo Catalyst