Catálogo de publicaciones - libros

Nuclear magnetic resonance (NMR) spectroscopy has long been applied to the detection and quantification of plant and fruit extracts, often after some kind of separation/purification treatment for simplification and concentration of the system. The challenge remaining regards the application of NMR to heterogeneous and intact systems, so that full use of the non-invasiveness of NMR is made. Tackling physically inhomogeneous materials such as intact fruits and vegetables or some of their components (e.g. skin, pulp, pips) must take into account that different physical phases are characterized by distinct magnetic susceptibilities and that anisotropic effects such as chemical shift anisotropy and internuclear dipolar interactions play increasingly important roles in semisolid and solid materials. The NMR lines in the spectrum are broadened to an extent related to the strength of the effect. Magnetic susceptibility effects may lead to hundreds of Hz to kHz line broadening, depending on the nature of the phases, whereas dipolar interactions may broaden NMR signals up to tens or even hundreds of kHz, in the extreme case of rigid solids. In order to recover spectral resolution, magic angle spinning (MAS) is employed. This consists of spinning the rotor containing the sample around an axis making an angle of 54°44′ with the direction of the external magnetic field B _0. Detailed descriptions of this procedure and its principles may be found elsewhere [1,2]. A necessary condition for MAS to result in satisfactory line narrowing is, however, that the spinning speed is at least of the same order of magnitude as the line broadening interaction. In the case of intact fruits, resolved spectra of tissues are typically achievable at spinning rates <1-2 kHz, whereas harder parts or components will require faster rates. In some cases, plasticization or swelling of the materials may be required to enable a resolved spectrum to be obtained, using technically achievable spinning rates, without risking mechanical damage of the sample (usually maximum rates used are 4-6 kHz). The designation of high-resolution (HR) MAS has typically been applied to the observation of the _1H nucleus under specific conditions (described below) that allow resolution improvement relative to the standard MAS method. However, due to the very limited applications of _1H HR-MAS to fruits and vegetables, the present text will also review the applications of standard _1H and _13C MAS to those samples.

High-resolution solid-state NMR has found widespread applications in various fields of agricultural and food sciences [1-6]. The technical improvements of high-resolution solid-state NMR have been tremendous for the last 10 years; the sensitivity and the design of probes, the setting and the control of spinning rates at the magic-angle spinning (MAS), the decoupling power strength, the decoupling schemes, and the large number of available pulse sequences are certainly among the most significant technical improvements [7,8].

Bread making is a process consisting of four basic steps, mixing, kneading, proofing, and baking. Although each step contributes to the set up of the aerated structure of bread, the baking process of the dough foam is the ultimate and more impressive part of the bread making process. The final open sponge structure with interconnected cells results from more or less simultaneous physical, chemical, and biochemical changes during baking such as volume expansion, evaporation of water, protein denaturation, starch gelatinization, and formation of a open network of pores.

Most MRI applications to dairy products are based on the sensitivity of relaxation time to chemical composition and changes in structure. It should therefore be borne in mind that for many dairy products the NMR relaxation signal is multi-exponential. This behavior is mainly explained by the chemical composition, since protons from water, fat, and protein contribute to the signal. Moreover, two phases can be observed according to the measurement temperature, i.e the liquid and solid fat. All these aspects complicate the interpretation of NMR relaxation time parameters and contrast in the MR image. Nevertheless, as the spin—spin relaxation time of protein protons and crystallized fat protons are very short, the image intensity is only dependent on the relaxation of the liquid phase, i.e the liquid fat proton and the water protons. Of course, if MRI acquisitions are performed at low temperatures where most of the fat is crystallized, the gray level intensity will describe only the water protons. This explains why most work has been done on low fat products or at low temperatures, or in high fat products such as cream and cheese, although some methods have been proposed to suppress water and fat signals selectively.

The flavor and the structure of crumb are the main factors in consumer acceptance of bakery products. The final structure depends to a significant extent on the creation and control of gas bubble structures in the unbaked matrix, and the retention of these gas bubbles in a suitable form until the matrix becomes set or baked. Bread making may be viewed as a series of aeration stages: bubbles are incorporated during mixing, the bubbles are inflated with carbon dioxide during proving, and the aerated structure is modified and set by baking. For greater convenience to craft bakers and greater compliance in logistics management, the immemorial continued process of proving and baking, has been sophisticated by introducing intermediary steps such as retarding techniques by cooling yeasted doughs, deep freezing of unproven fermented doughs, or part-baked bread in centralized production plants (Figure 1). The new steps have contributed to make more complex the physical and chemical changes that occur inside the dough during processing. Bread making quality of frozen green dough is affected by the degradation of both yeast activity and dough rheology. The water crystallization process plays a major role in such degradations. The rate at which ice forms with lowering temperature, the microscopic structure of ice crystals as well as its evolution during cold storage are known as parameters of importance. As invasive measurements may provoke dough collapse, the characterization of the expansion process has been reduced to a number of global volumetric parameters. In the last decades, sophisticated measurements and macroscopic invasive techniques have allowed a better understanding of these changes. If the resolution of confocal microscopy is efficient to yield the number of bubbles and the porosity of dough, only bubbles close to the dough surface can be observed. If the X-ray tomography and the MRI technique both have the potential to investigate the porosity in all parts of the dough, only MRI offers the possibility to measure simultaneously porosity and to localize ice gradients during the thawing-proving process. Actually MRI parameters characteristic of the dough, i.e. transversal T ^2 and longitudinal T ^1 relaxation times, and apparent proton densities, are sensitive to structural changes such as water crystallization, but also moisture modifications, expansion process, and temperature changes in the matrix. The codification of the MRI images based on the application of magnetic gradients in each space direction before spin or gradient echo scanning, combined to the rapid relaxation of water protons in dough (e.g. T ^2 is comprised between 0.5 and 35 ms depending on temperature) confined the technique to the partial observation of the more mobile protons: a more or less important fraction of the protons of liquid water and a fraction of the hydroxyl protons of macromolecules in interaction with water. Large reduction in the MRI signal is associated, on the one hand, to the crystallization from up to 50–60% of the water during freezing of green dough and, on the other hand, to the expansion during proving with the incorporation of up to 60–70% of gas in bubbles. This involves to pay attention to maintain exploitable signal-to-noise ratio all along the processes, by finding the best compromise in the choice of the spatial resolution, the echo time and the number of excitations for signal averaging. After that, the complex relationship between gray levels in the dough images and physical state of the different parts of the dough requires large investigations to quantify the precise contribution of each phenomenon. This review presents the state of art of the performances of the MRI technique to visualize and quantify the phenomena occurring during the making of dough (freezing and proving) and to help their complex modeling.

Most foods (initial/final) and food components are disperse multiphase systems (suspensions, foams, emulsions, porous/dry/wet bulk solids). Aspects of the material properties essential for both consumer and producer are the structure (different length-scales) and flow behavior. The flow behavior is important for the design of process unit operations and sensory behavior of the final food product. The structure is correlated with various macroscopic properties like the storage/ageing behavior, phase separation (syneresis, sedimentation) [1,2], and the filtration/deformation/fiow/sensorial behavior. Correlations between the evolution of the stress and porosity, agglomeration, and flow-type are established for flowing foods (starch solutions [3], wheat dough [4]). It is, therefore, plausible to assume that the correlation between the structure and the rheology is generally valid. Materials with volume fractions of the disperse phase higher than 5 vol.% are usually opaque [5]. MRI offers the possibility to perform measurements non-invasively, non-destructively, and highly selective ( in situ , online) in order to study the structure or trace structural changes or inner transport processes in different phases/states of aggregation (without preparations) during production, storage, transportation, and consumption (temporal, spatial, chemical resolved). Samples may be studied several times under identical or changed conditions, which is a great advantage (time saving, higher accuracy, interior surface) compared to preparative microscopy. The objective of MRI in food process engineering is to develop and check appropriate models and describe the system for a reliable process design, quality control, and process control, in order to forgo time-intensive trial and -error. Foods have often to be built up step by step from simple ingredients to complex systems in order (i) to study effects of single constituents and (ii) to quantify interactions between constituents. On the basis of the knowledge of the involved phenomena, new products/apparatus with improved process conditions can be developed economically [6].

In food process engineering, the handling, pumping, extrusion, or mixing of highly concentrated disperse systems is of importance for initial, intermediate, or final products. Examples are suspensions (beer mashes, chocolate, dough), foams (protein-, carbohydrate-based), emulsions, porous solids (baked products), and dry/wet bulk solids (malt, sugar). Although plenty of information concerning the rheology of multiphase system is available, the modeling of the flow behavior, structure of the produced system, and its stability cannot be considered as totally completed. In spite of its great importance, the determination of viscosities of disperse systems is difficult in principle [1,2]. Defined velocity profiles in rheometers (so-called viscometric flows), the homogeneity of the sample during the experiment, and applicability of continuous mechanics are usually prerequisites for rheometric studies. In disperse multiphase systems, demixing (phase separation: sedimentation, creaming, wall layers), scale effects [3], preparation, time effects (crushing, de-/agglomeration), tearing apart (fat, ointments), and wall slip cause problems when studying the flow behavior and determining flow functions of disperse systems with conventional rheometric methods.

Several types of NMR experiments result in data that may be described as a sum of exponential functions ( T ^1 weighted signals, T ^2 weighted signals, and gradient weighted signals). In order to be able to understand and interpret these data, it is imperative to be able to model these with a robust and precise method which is why the numerical scientists have worked for centuries with iterative non-linear least squares curve fitting methods for resolving underlying mono-exponential functions. Recently, a new non-iterative method was proposed by Windig and Antalek [1] called direct exponential curve resolution algorithm (DECRA). The new method has several potential advantages, but first of all it is non-iterative, accurate, and rapid. In this chapter, a more general version of the DECRA method called Slicing is outlined and applied to a simple test case measuring fat in meat by NMR relaxometry.

Oxidative reactions in food that involve lipids are, together with microbial growth, among the most important causes of spoilage of food and affect both shelf life and the quality of foods. The oxidation of unsaturated lipids in foods is observed as rancidity, while oxidation of proteins may result in other off-flavors and in some cases as a change in texture. The oxidation of vitamins and polyunsaturated fatty acids reduce the nutritional value and form potentially toxic oxidation products, while oxidation of food pigments leads to discoloration. Radicals are intermediates in many of these reactions, and the detection of radicals by electron spin resonance, ESR, provides therefore an excellent way to study early stages of these destructive reactions [1-3]. The study of radicals by ESR can moreover give important mechanistic information, which is useful for designing protective measures against oxidation in food. The detection of radicals by ESR has also recently also been used to predict the oxidative stability of foods, and further to quantify important quality aspects of food.

A relatively new approach in NMR is the relaxing of restrictions on the magnetic field homogeneity for the static field B ^0 as well as for the radio frequency field B ^1, known as single-sided or unilateral NMR. This realization of an NMR device implies cost reduction of the equipment as well as the possibility to non-destructively apply NMR techniques to samples larger than the probe. However, this approach has also physical consequences. The magnetic field gradients are in an order of magnitude where almost every pulse becomes selective thus leading to spatial selectivity which is given by the Larmor frequency and the bandwidth of the tank circuit. In addition, these single-sided NMR devices have an inherent sensitivity to diffusive or convective processes as the field gradients are large. For example, the device used for the measurements presented here exhibits B ^0 field gradients in the order of 10–20 T/m, leading to an full width at half maximum, FWHM slice thickness of about 2 mm with a quality factor of the circuit of about 70. The B ^0( r ) can be constructed such that shape and size of the sensitive volume corresponds to the sample to be measured and is therefore to be tailored to the desired properties. In case of homogeneous food samples, the sensitive volume should be chosen as large as possible, whereas small and sharp volumes are preferred in examinations of local properties. Examples are the investigations of skin cancer and of fat content in living salmon [1].