Molecular Solution Behaviour of an Intermediate Biofuel Feedstock: Acetone–Butanol–Ethanol (ABE)

Abstract

Mixtures of acetone, butanol, and ethanol (ABE) are common intermediate products in the production of biofuels via biomass fermentation. Their separation to yield, for example, bio-butanol, is still difficult due to the lack of a fundamental understanding of these mixtures at the molecular level. In order to bridge this gap, a detailed analysis of characteristic features of the vibrational spectrum is carried out. A systematic study of the binary solutions of acetone with ethanol and butanol does not only reveal a universal behaviour at the molecular level when acetone is mixed with short-chain alcohols, it also shows that the phenomena at a length scale between the molecules and in the macroscopic solution need to be taken into account to understand the structure-property relationships. The size of self-associated molecule clusters seems to determine whether or not a system exhibits an azeotrope. When a second alcohol is added to an acetone/alcohol solution, no additional non-idealities are induced, which is advantageous for modelling ternary ABE mixtures and for improving their processing in the production of biofuels.

Unsupervised Screening of Vibrational Spectra by Principal Component Analysis for Identifying Molecular Clusters

“…So despite the small variance of PC3 in the initial data set it seems to have a physical meaning, namely it points towards the ethanol‐butanol dimers. Interestingly, the sole analysis of the excess spectra in a previous article did not allow distinguishing between butanol‐ethanol dimers and butanol‐butanol and ethanol‐ethanol clusters . Hence, the application of PCA reveals deeper insights in the cluster formation compared to the sole analysis of excess data even for ideal systems.…”

“…For further details on the data acquisition and treatment, for example, how the excess spectra were obtained, see the references. [12,17]…”

“…16 scans were averaged to obtain an appropriate signal‐to‐noise ratio. For further details on the data acquisition and treatment, for example, how the excess spectra were obtained, see the references …”

“…Moreover,t he excess spectrumw ill essentially be zero in an ideal solution.S uch ab ehavior can be observedi nm ixtures of chemically similar compounds, for example, mixtures of primary alcohols. [12] ApplyingaPCA on the IR or excess IR spectra of such mixtures does not provide much information. As an example,binarym ixtures of ethanola nd butanol were investigated.…”

“…Interestingly,t he sole analysiso ft he excess spectra in ap revious article did not allow distinguishing between butanol-ethanol dimers and butanol-butanol and ethanol-ethanol clusters. [12] Hence, the application of PCA reveals deeper insights in the clusterf ormationc ompared to the sole analysis of excessd ata even for ideal systems. In this context,m ultivariate curve resolution (MCR) would probably have yieldeds imilar results.…”