Soft Independent Modelling of Class Analogy applied to infrared spectroscopy for rapid discrimination between hardwood and softwood

Published Date
Energy
15 December 2016, Vol.117:251–258, doi:10.1016/j.energy.2016.10.092

• Author 

• D. Duca
• M. Mancini
• G. Rossini
• C. Mengarelli
• E. Foppa Pedretti
• G. Toscano
• A. Pizzi ^,

• Università Politecnica delle Marche - Agricultural, Food and Environmental Sciences, Via Brecce Bianche, 60131 Ancona, Italy

Received 15 October 2015. Revised 14 September 2016. Accepted 21 October 2016. Available online 28 October 2016. 

Highlights

• •
  Quality parameters such as origin and source are required by pellet standards.
• •
  The study aims to discriminate hardwoods and softwoods.
• •
  FTIR spectroscopy coupled with SIMCA has been evaluated as a suitable tool.
• •
  110 wood samples belonging to 12 species and 4 blends were analysed.
• •
  The method shows good performance and could be useful to verify standard compliance.

Abstract

European policy promotes renewable energies and sets specific targets. Solid biofuels can play a significant role and the quality is an important aspect to be checked. Quality parameters such as origin and source are also required by specific biofuel standards. Therefore it could be useful to develop a rapid and cheap tool to distinguish between hardwoods and softwoods especially in unstructured, milled or densified wood in order to check the compliance of producer's declarations. Compared to other analytical methods, infrared spectroscopy is fast, non-destructive and low cost. In this study Fourier transform infrared (FTIR) spectroscopy coupled with Soft Independent Modelling of Class Analogy (SIMCA) has been evaluated as a method for discrimination purpose. A large dataset of 110 wood samples belonging to 12 species were analysed. In addition 4 blends were also analysed to test the discrimination performance of the tool. FTIR-SIMCA has correctly classified 93% of hardwood samples and 100% of softwood samples at high significance level. Furthermore, 100% of tested blends were associated with no class, showing a discrimination ability of the technique in recognizing blends from pure material. This method could be useful to verify the compliance of producer declarations about wood origin and source.

Keywords

MIR

Pellet

Woody biomass

SIMCA

Multivariate analysis

Biomass blends

---

 Table 1

Table 1.

Fig. 1.

 Table 2

Table 2.

Fig. 2.

Fig. 3.

 Table 3

Table 3.

Fig. 4.

Fig. 5.

Fig. 6.

 Table 4

Table 4.

• ∗ 
  Corresponding author.

For further details log on website :
http://www.sciencedirect.com/science/article/pii/S0360544216315377