Elsevier

Food Chemistry

Volume 173, 15 April 2015, Pages 1220-1227
Food Chemistry

Application of Rapid Visco Analyser (RVA) viscograms and chemometrics for maize hardness characterisation

https://doi.org/10.1016/j.foodchem.2014.10.149Get rights and content

Highlights

  • This study provides a method to describe maize hardness and milling quality.

  • Rapid Visco Analyser data was used in combination with chemometrics.

  • A LW-PLS2 regression model was used.

  • Maize hardness could be predicted with success.

Abstract

It has been established in this study that the Rapid Visco Analyser (RVA) can describe maize hardness, irrespective of the RVA profile, when used in association with appropriate multivariate data analysis techniques. Therefore, the RVA can complement or replace current and/or conventional methods as a hardness descriptor. Hardness modelling based on RVA viscograms was carried out using seven conventional hardness methods (hectoliter mass (HLM), hundred kernel mass (HKM), particle size index (PSI), percentage vitreous endosperm (%VE), protein content, percentage chop (%chop) and near infrared (NIR) spectroscopy) as references and three different RVA profiles (hard, soft and standard) as predictors. An approach using locally weighted partial least squares (LW-PLS) was followed to build the regression models. The resulted prediction errors (root mean square error of cross-validation (RMSECV) and root mean square error of prediction (RMSEP)) for the quantification of hardness values were always lower or in the same order of the laboratory error of the reference method.

Introduction

Maize hardness is an important quality characteristic for the dry-milling industry. In South Africa, the dry-milling industry is significant as maize is the largest crop produced, of which two-thirds (ca. 4 million tons per annum) are processed into maize meal used to make porridge (SAGIS, 2013). For its definition, numerous aspects have to be considered. Even with careful consideration, the interpretation and measurement of maize hardness can still be confusing and needs further investigation. Nonetheless, the hardness of maize is important when determining the processing settings, such as for dry-milling. Similar to other grains, maize kernel hardness is principally a genetic characteristic (Johnson & Russell, 1982), although environmental influences (Hamilton, Hamilton, Johnson, & Mitchell, 1951) and external factors such as postharvest handling (Peplinski, Paulsen, Anderson, & Kwolek, 1989) will also affect the hardness. Maize (Zea mays L.) is anatomically made up of two types of endosperm, i.e. a harder (vitreous) endosperm situated to the outside of the kernel, and a softer (floury) endosperm found in the center of the kernel (Paiva et al., 1991, Watson, 1987). It is known that hard kernels are favoured by industry as hard maize produces greater yield and a higher quality meals and grits than soft maize (Lee et al., 2007).

There are many methods available to determine maize hardness, as extensively reviewed by Fox and Manley (2009). Yet, it is not clear which method best describes hardness. It is even more difficult to decide on a method that best describes milling quality. In this study, the samples were ranked according to their milling performance as measured during the actual milling process, although on pilot plant-scale. The outcome of good milling is indicated by a small percentage of hominy chop. Hominy chop (comprising the pericarp, tip cap, germ and some endosperm) is of lesser value than maize meal and grits and predominantly used as animal feed. Maize that mill poorly delivers a larger percentage chop (%chop) as soft endosperm is also included into the chop. Percentage chop is therefore used as an indication of the milling quality of maize, although it is not a recognised hardness measurement method as such. In this study, the relationship of %chop to other conventional hardness methods was investigated. These conventional methods were chosen to include many different descriptors, i.e. density, size and soundness (hectoliter mass and hundred kernel mass), particle size or breakage susceptibility (particle size index and near infrared absorbance) and quality properties (protein content and near infrared hyperspectral imaging).

Great financial losses experienced in the Australian wheat industry in the 1980’s lead to the development of the Rapid Visco Analyser (RVA) (Ross, Walker, Booth, Orth, & Wrigley, 1987). This viscometric method has since found meaningful applications in a vast range of applications, especially in grain science due to the large amount of starch present in cereals (Agu et al., 2006, Doublier et al., 1987). The RVA measures the viscosity developed with hydration and subsequent gelatinisation of starch granules during heating and stirring in excess water (Almeida-Dominguez, Suhendro, & Rooney, 1997). It has been reported that the RVA can be used to quantify maize hardness differences between maize hybrids Ji et al., 2003, Sandhu and Singh, 2007, Seetharaman et al., 2001, Yamin et al., 1999. This was due to hard maize producing mainly coarse particles when milled, and soft maize smaller particles (Almeida-Dominguez et al., 1997). Coarse particles have slower water diffusion, limited swelling of the starch granules and slow viscosity development (Narváez-González et al., 2006, Sahai et al., 2001, Sahai et al., 2001). Smaller particles have bigger surface areas that result in better and more rapid hydration, thus better gelatinisation and higher viscosity (Almeida-Dominguez et al., 1997). Furthermore, hard kernels show a more prominent protein-to-starch adhesion effect compared to soft kernels (Almeida-Dominguez et al., 1997). The protein matrix of vitreous endosperm is thicker than that of floury endosperm (Wang & Eckhoff, 2000), and forms a barrier that slows hydration and gelatinisation (Almeida-Dominguez et al., 1997, Narváez-González et al., 2006).

Based on these considerations, determination of the usefulness of the RVA as a hardness descriptor was a key concern in the current study. With this aim, RVA viscograms were recorded on different maize samples which were also quantified and characterised for hardness by the seven conventional (reference) methods. The objective was to define whether it was possible to obtain information about properties (such as density, breakage susceptibility and protein content), that are commonly associated with hardness, from RVA curves obtained through single measurements. In order to analyse such a complex data set, the use of appropriate chemometrics techniques was required. At first, principal component analysis (PCA) was used to achieve a better understanding of the relations among the seven reference methods. Subsequently, the use of a non-linear regression technique, i.e. locally weighted partial least squares (LW-PLS) regression (Bevilacqua et al., 2012, Centner and Massart, 1998) was used to build a regression model to predict maize hardness from the RVA curves. This was done for all seven reference methods individually.

Section snippets

Maize samples

Nineteen pure hybrids of South African white maize, originating from maize breeding trials, were used. These hybrids, kindly supplied by PANNAR Seeds (Greytown, South Africa) came from three localities (Greytown, Delmas and Klerksdorp, South Africa) and two plantings in 2012 (early and late). A local farmer from Schweizer-Reneke also supplied some of the samples, resulting in a total of 49 samples. Before being milled, the samples were stored at ambient temperature in sealed plastic containers.

Pasting properties (acquisition of RVA data)

Results and discussion

Descriptive statistics of the seven hardness related reference methods are summarised in Table S1.

Conclusion

This study attempted to provide a solution to the problem of selecting the most appropriate method that best describe maize hardness milling quality. The different conventional methods were all shown to contribute towards describing hardness with %chop being the most appropriate to determine milling quality. The RVA was shown as being useful, not only for describing pasting properties of starch, but also to characterise maize hardness. Using LW-PLS2 regression, calibration models were developed

Acknowledgements

This work is based on research supported in part by the National Research Foundation of South Africa (Grant specific unique reference number (UID) 76641, 70863 and 83974). Gratitude to Ayanda Myende for assistance with RVA analysis and Prof Martin Kidd for performing univariate statistical analysis. The authors acknowledged Sasko, a division of Pioneer Foods (Pty) Ltd (Paarl, South Africa) for the use of their research and development facilities, supplying analytical reference results

References (33)

  • V. Centner et al.

    Optimization in locally weighted regression

    Analytical Chemistry

    (1998)
  • G. Downey et al.

    Wheat trading in the Republic of Ireland: The utility of a hardness index derived by near infrared reflectance spectroscopy

    Journal of the Science of Food and Agriculture

    (1986)
  • Emvula, S. (2012). Assessment of hectolitre mass (HLM) equipment and HLM measurements of oats. Masters. South Africa:...
  • G. Fox et al.

    Hardness methods for testing maize kernels

    Journal of Agricultural and Food Chemistry

    (2009)
  • P. Geladi et al.

    Linearization and scatter-correction for near-infrared reflectance spectra of meat

    Applied Spectroscopy

    (1985)
  • T. Hamilton et al.

    The dependence of the physical and chemical composition of the corn kernel on soil fertility and cropping system

    Cereal Chemistry

    (1951)
  • Cited by (19)

    • Effect of oven and forced convection continuous tumble (FCCT) roasting on the microstructure and dry milling properties of white maize

      2017, Innovative Food Science and Emerging Technologies
      Citation Excerpt :

      The HLM (in kg hL− 1) of the samples was determined using a German Kern 220/222 Grain Sampler (KERN & SOHN GmbH, Balingen-Frommern, Germany). This test was performed according to the method described earlier (Guelpa, Bevilacqua et al., 2015). Maize was degermed using a pilot plant scale degermer and subjected to the milling process.

    • A high-throughput X-ray micro-computed tomography (μCT) approach for measuring single kernel maize (Zea mays L.) volumes and densities

      2016, Journal of Cereal Science
      Citation Excerpt :

      Milling quality was described in the direction of PC 1. This was in accordance with the hardness (hard and soft kernels) study of Guelpa et al. (2015b). The loadings indicated the relationship between the variables, and in particular, it could be observed that a stronger correlation existed between V:F, EKD, VEV, VED, FED and EKV.

    View all citing articles on Scopus
    View full text