Test method for low level detection of biodiesel in diesel using the Agilent 5500t FTIR spectrometer

Importance of the Topic

The reliable detection of low‐level biodiesel (fatty acid methyl esters, FAME) in petroleum diesel is critical for industries that require strict fuel specifications. While blends up to 5 % FAME comply with ASTM D975 without disclosure, certain users—power generation, precision engines, and sensitive equipment operators—need to monitor and minimize trace biodiesel contamination to avoid performance or warranty issues.

Objectives and Study Overview

This application note presents an enhanced FTIR spectroscopic method that combines the high sensitivity of EN 14078 transmission sampling with the robust multivariate calibration of ASTM D7371. The goal is to accurately quantify biodiesel in diesel at levels from 0.025 % to 20 % FAME. The method was evaluated through cross validation and a blind round‐robin study, demonstrating superior accuracy, especially at low contamination levels.

Methodology and Instrumentation

The core analytical approach uses Fourier Transform Infrared (FTIR) spectroscopy with a controlled 100 µm transmission pathlength. Key steps include:

• Sample preparation: A single drop of fuel is sandwiched between rotating sapphire windows to achieve reproducible optical pathlength and ease of cleaning.
• Spectral acquisition: Mid‐IR spectra are recorded in the carbonyl stretching region (approximately 1950–1720 cm⁻¹) and additional ester bands (1327–1119 cm⁻¹) for the high‐range model.
• Data preprocessing: Mean centering, baseline correction, and thickness correction are applied before calibration.
• Multivariate calibration: Partial least squares (PLS) models are developed in three concentration ranges—0.025–1 %, 1–10 %, and 10–25 % FAME—automatically selected by control software based on the sample’s apparent concentration.

Used Instrumentation

The method employs the Agilent 5500t FTIR spectrometer equipped with a patented transmission sampling interface. This modular accessory provides a 100 µm pathlength with reproducibility better than 0.2 µm while maintaining the simplicity of ATR sampling for viscous fuels.

Main Results and Discussion

Calibration performance for each PLS model was assessed by leave‐one‐out cross validation and external validation sets:

• Range 0.025–1 %: SECV = 0.0016 %, R² = 0.9999, average validation error = 1.37 %
• Range 1–10 %: SECV = 0.0164 %, R² = 0.9999, average validation error = 0.06 %
• Range 10–20 %: SECV = 0.04 %, R² = 0.9999, average validation error = 0.57 %

In a blind round‐robin comparison with five other laboratories, the Agilent method achieved a total average relative error of 2.1 % across 2–20 % FAME samples, and only 1.1 % error at low‐level concentrations—outperforming ASTM D7371 ATR‐based approaches.

Benefits and Practical Applications

• High sensitivity for trace biodiesel down to 0.025 % ensures compliance and quality control in sensitive applications.
• Rapid analysis (<5 minutes per sample) with minimal sample handling.
• Automated model selection simplifies operator workflow and reduces potential errors.
• Reproducible pathlength control and easy cleaning minimize maintenance.

Future Trends and Potential Applications

Advancements in portable FTIR devices may enable real‐time, on‐site fuel monitoring. Combining multi‐attribute spectral data with machine learning could further enhance specificity for different biodiesel feedstocks or contaminant profiling. Integration with IoT and cloud‐based analytics offers the potential for fleet‐wide fuel quality surveillance.

Conclusion

By uniting the transmission sampling sensitivity of EN 14078 with the multivariate PLS algorithm and standards of ASTM D7371, the Agilent 5500t FTIR method delivers precise, low‐level biodiesel quantification from 0.025 % to 20 % FAME. With demonstrated superior accuracy and ease of use, this approach addresses the needs of users requiring stringent fuel purity verification.

Reference

John Seelenbinder and Frank Higgins, Agilent Technologies Application Note 5990-7804EN, published May 1, 2011.