Biodiesel (FAME) Analysis by FT-IR

Significance of the topic

Biodiesel (fatty acid methyl esters, FAME) is growing in importance as an alternative or additive to petroleum diesel due to fuel-cost pressures, environmental drivers and policy incentives. FAME improves diesel lubricity (even at low blend levels), reduces sulfur and aromatic emissions, and can be blended at defined percentages (B# notation). Reliable, rapid, and reproducible analytical methods are essential both for producers (process control, removal of glycerol, completion of transesterification) and for regulatory or terminal testing (verification of labeled blend levels across a wide FAME range from B2 to B100). FT-IR with ATR sampling provides a practical solution for many of these needs.

Goals and overview of the study

This application note demonstrates using mid-infrared FT-IR spectroscopy with an ATR accessory and chemometric calibration (partial least squares, PLS) to quantify FAME in diesel/biodiesel blends. The objectives were to establish a robust, rapid measurement workflow that: (1) identifies spectral markers for FAME, (2) defines instrument/ATR configuration choices for different concentration ranges, (3) builds PLS calibrations that account for matrix variability in petroleum diesel stocks, and (4) explores calibration transfer between instruments.

Used instrumentation

• Nicolet 380 FT-IR spectrometer with KBr beamsplitter and a DTGS detector.
• Smart ARK attenuated total reflectance (ATR) accessory with interchangeable ZnSe top plates (45° and 60°) to adjust effective path length.
• Data collection parameters: 32 scans, 4 cm⁻¹ resolution, ~40 s per spectrum.
• Sample handling: ~0.4 mL required; crystal cleaned with acetone-toluene-methanol mixture or by use of the next sample; tri-solvent cleaning before background collection.
• Software: OMNIC for spectral acquisition and TQ Analyst for chemometrics (PLS). Thermo Scientific ACLS algorithm used for calibration transfer (inoculation) between instruments.

Methodology

Mid-IR spectra of diesel and FAME were acquired by ATR. Key FAME absorptions are the ester carbonyl C=O near 1750 cm⁻¹ and C–O stretching bands around 1170–1200 cm⁻¹; the 1750 cm⁻¹ band is generally free of interference from petroleum diesel, while the 1170–1200 cm⁻¹ region can overlap with variable petroleum signals. PLS regression models were developed following emerging ASTM guidance to correlate spectral variation with FAME concentration. Two ATR plate geometries were evaluated: the 45° ZnSe plate (stronger signals, suitable for low-percentage blends) and the 60° ZnSe plate (reduced effective path length, avoids detector saturation at high FAME levels). The absorbance linearity criterion (instrument limit ~1.2 AU) guided plate selection. Calibration robustness was improved by including representative petroleum diesel variations; a small number of additional standards from a different petroleum base (“inoculation”) markedly improved prediction accuracy. Calibration transfer between spectrometers used the ACLS algorithm to correct offsets prior to applying a PLS model on a second instrument.

Main results and discussion

• Spectral markers: Strong, diagnostic ester peaks at 1750 cm⁻¹ (C=O) and 1170–1200 cm⁻¹ (C–O) underpin the quantitative approach; the 1750 cm⁻¹ region is especially valuable due to minimal petroleum interference.
• ATR plate performance: The 45° plate produced larger ester absorbances and is preferable for low-FAME ranges, but FAME absorbance exceeded the recommended linear limit (~1.2 AU) above ~B30; the 60° plate reduced absorbance (example ~0.8 AU for neat FAME) allowing use across higher concentrations without saturation. Plate swapping is reproducible and does not require optical realignment or full recalibration.
• Calibration accuracy: PLS models built with appropriate standards produced strong predictive performance. Reported prediction errors were approximately ±0.07 FAME percentage (low-range/45° plate) and ±0.1 percentage (high-range/60° plate) for unknown samples.
• Matrix variability: Initial calibrations built using a single petroleum base gave poor predictions when applied to blends made with a different diesel stock. Adding only three representative standards from the alternate petroleum base (“inoculation”) dramatically improved predicted values (examples: actual 20% FAME predicted 14.36% initially, corrected to 20.01% after inoculation).
• Calibration transfer: ACLS algorithm implementation removed small offsets between instruments, enabling transfer of calibrations between different spectrometers after limited local inoculation—consistent with practices used in FT-NIR.
• Extensibility: While the study focused on total FAME, the authors note FT-IR with appropriate calibrations can quantify other analytes relevant to biodiesel quality control, such as free fatty acids and glycerol.

Benefits and practical applications

• Speed and simplicity: ATR-FTIR provides rapid results (~40 s per spectrum) with minimal sample preparation and small sample volumes (~0.4 mL), suitable for routine QC at production sites and terminals.
• Robust quantitation: Targeting the 1750 cm⁻¹ ester band and using chemometrics yields precise, reproducible FAME determinations across typical blend ranges.
• Adaptability: Interchangeable ATR plates permit optimization for different FAME concentrations without complex instrument adjustments.
• Calibration strategies: Inclusion of representative petroleum matrices and use of inoculation techniques enable reliable predictions across regional diesel variations; ACLS supports calibration transfer across instruments for multi-site implementations.
• Regulatory and operational value: Method aligns with ASTM committee recommendations for PLS-based analyses and can support both producer process control (B100 quality) and regulatory/terminal verification of blend labels (B2–B100).

Future trends and potential applications

• Broader multi-analyte methods: Development of generalized FT-IR chemometric models that quantify FAME, free fatty acids, glycerol and other impurities from the same spectrum.
• Improved calibration robustness: Expanded calibration sets covering wider petroleum feedstock variability, seasons, and regional fuels to minimize inoculation needs.
• Automation and inline monitoring: Integration of ATR-FTIR into process analytical technology (PAT) for near-real-time monitoring of transesterification and wash/purification steps.
• Calibration transfer and networking: Standardized transfer algorithms and cloud-shared calibration libraries to streamline multi-site deployments and instrument upgrades.
• Portable/field instruments and FT-NIR complementarity: Development of field-deployable infrared platforms and comparative workflows using FT-NIR for rapid screening with FT-IR confirmation.

Conclusion

ATR-FTIR combined with PLS chemometrics is an effective, rapid, and practical method for quantifying FAME in diesel/biodiesel blends. Key advantages include minimal sample volume, fast acquisition, straightforward sample handling, and the ability to tune sensitivity via interchangeable ATR plates. The method requires representative calibration standards to capture matrix variability in petroleum diesels, but limited inoculation of a calibration set or application of instrument-correction algorithms enables robust predictions and transfer between instruments. With appropriate calibration development, FT-IR can be extended beyond FAME quantitation to other quality attributes relevant to biodiesel production and regulatory compliance.

References

• Bradley M. Biodiesel (FAME) Analysis by FT-IR, Application Note 51258, Thermo Fisher Scientific, 2007.