Read an Excerpt
NMR Spectroscopy in Food Analysis
By Apostolos Spyros, Photis Dais The Royal Society of Chemistry
Copyright © 2013 Apostolos Spyros and Photis Dais
All rights reserved.
ISBN: 978-1-84973-533-9
CHAPTER 1
Introduction
Nuclear magnetic resonance (NMR) spectroscopy is an effective analytical technique, which has been used systematically in food analysis and authentication in recent years. Its origin is traced back to 1946 when two groups of scientists at Harvard University (Purcell, Torrey and Pound) and at Standford University (Bloch, Hansen, and Packard), working independently, observed proton resonance signals from paraffin wax and water, respectively. For their discovery, Purcell and Bloch were jointly awarded the Nobel Prize in Physics in 1952.
The first application of NMR in food science dates back to 1957 when low-resolution NMR measured moisture in foods. Consistent and widespread application of NMR in food science started in the 1980's mainly due to deficiencies in instrumentation and the complexity of food matrices. Since then, an explosive publication of research and review articles dealing with NMR applications in food science has appeared in scientific journals and several books. Figure 1.1 shows diagrammatically the explosion of publications in food science after 1988.
Also, numerous oral and written communications have been presented in domestic and international conferences. In particular, an International Conference on Application of Magnetic Resonance in Food Science is held in Europe every two years. This conference started in 1992 and gave the opportunity for scientists worldwide to present new applications of NMR in food science and technology. It is worth mentioning that NMR methods have been approved as official methods by the European Union (e.g. detection of wine fraud). There are several reasons for this development: (a) the increasing sophistication and the user-friendly NMR instrumentation; (b) the increasing need of the food industry to understand innovate its products and processes; and (c) the necessity for the development of new and more effective analytical techniques for the quality control and authentication of foods and thereby the reinforcement of pertinent legislation.
Foods are very complex and highly heterogeneous systems comprising a large number of chemical compounds, the composition of which varies considerably under certain circumstances (e.g. agronomical or slaughter practices, industrial processes, storage, maturation, etc.). To this direction, one-dimensional (1D) liquid or solid-state high-resolution NMR spectroscopy can provide in a single experiment a wealth of structural and quantitative information in the form of the NMR parameters, namely chemical shifts, coupling constants and signal intensities. For the same sample the researcher can choose different nuclei, such as 1H,13C,31P,19F — to mention the most popular nuclei — that allow the study of food samples under different perspectives and to extract the maximum information about its natural or industrial condition. These experiments need no separation of the various food components and no serious sample pre-treatment. Moreover, NMR spectroscopy is sensitive to dynamics, which allows differentiation between molecules or groups of molecules with different mobility through spin–lattice and/or spin–spin relaxation measurements.
In cases where the complexity of the food sample is so severe, causing extensive signal overlap in 1D spectra, the arsenal of NMR spectroscopy provides a large number of analytical techniques starting from the homonuclear and heteronuclear multi-dimensional NMR to its hyphenation with effective separation techniques, such as liquid chromatography (LC-NMR). In particular, two-dimensional (2D) NMR techniques, such as COSY, TOCSY, NOESY, HSQC, etc., based on the inherent 'communication' of nuclei with each other (through spin–spin and dipolar coupling), spread out the spectroscopic information in two dimensions unravelling hidden nuclear connectivities and facilitating the structural characterisation of the molecules in the food sample. Although NMR spectroscopy is not a destructive analytical technique and recovery of the analyte can be easily achieved after experimentation, industrial needs may require the examination of food products under different processing conditions by non-invasive means. Magnetic resonance imaging (MRI), used extensively in medicine, has been exploited in recent years in food analysis. The ability of MRI to show spatial resolution within the food product and the judicial application of MRI techniques allows the monitoring of the fate of certain molecules (e.g. water) and reveals various molecular interactions and changes in tissue structure that occur during food processing or storage (e.g. food freezing and thawing).
The combination of NMR spectroscopy with multivariate statistical methods provided an alternative possibility of analysing and maximising the information recovery from complex NMR spectral data of foods. This methodology, usually called metabonomics, does not necessarily require the identification of the individual signals in the spectrum as in quantitative NMR, but seeks to find subtle spectral features that can identify unequivocally the presence of metabolites or useful biomarkers. Pattern recognition techniques (supervised or unsupervised) can be used to map the NMR spectra of a large number of samples, and locate spectral fingerprints that reflect either metabolic changes or used to distinguish sample classes.
The disadvantage of the early days of NMR spectroscopy related to the low sensitivity and high cost of the analysis does not hold true for the NMR instrumentation of the present day. These drawbacks have been largely compensated by the development of modern hardware comprising strong magnetic fields up to 23.5T and cryogenic probes that allow easy detection of food components at the level of µg and even ng. Moreover, the progress in sophisticated software and innovations in automation allow the screening of a large number of samples (overnight run), reducing the experimental time to a few minutes even for the less sensitive nuclei (e.g.13C,15N).
In concluding this introductory chapter, we could add that it is not only the unique information that NMR provides, but also the versatility of methods, instruments and probes that make it an important tool for qualitative and quantitative analysis.
This book has been organised as follows: Chapter 1 is the book's introduction. Chapter 2 gives an account of the theory underlying the physical phenomenon of NMR and grouping the most useful NMR techniques to better understand the core principles, which appear in subsequent chapters. Since NMR is a well-documented spectroscopic technique and it is well described in several introductory and advanced books, this chapter will be kept to a minimum. Chapter 3 describes the NMR instrumentation in an attempt to familiarise the reader with the hardware and software components of modern high-resolution and solid-state NMR spectrometers, and their functions and automation. Relevant information about NMR spectrometers and the implementation of its components may help the reader to choose the right spectrometer and accessories for their needs. Also, this chapter includes useful information about the hardware systems and experimental designs to perform sophisticated experiments, such as (HP)LC-NMR, time-domain NMR, and high-throughput and on line NMR. Appropriate guidance for obtaining pure samples from various food matrices that are suitable for NMR experiments will be presented in Chapter 4, whereas the experimental conditions described in Chapter 5 may help the NMR user to choose the right input values for the critical parameters in the experimental setup in order to obtain the maximum possible information from the NMR experiment, and to perform quantitative analyses with high accuracy and precision. Chapter 6 presents a few aspects of the supervised and unsupervised pattern recognition statistical methods employed for data exploration, classification of food samples, and the build-up of calibration–prediction models giving special attention to NMR metabonomics. The applications of NMR spectroscopy and its specialties to different food systems are discussed in Chapters 7–11. A detailed presentation of the available NMR methodologies and techniques for each food category is provided, whereas practical guidance and tips for performing concrete experiments is afforded. Every chapter starts with a short abstract and ends with relevant bibliographic coverage.
CHAPTER 2
Theoretical Aspects
2.1 Nuclear Spins and Energy States
An atomic nucleus is a collection of protons and neutrons (nucleons) that possess a quantum mechanical property called spin, which is characterised by spin angular momentum. Spin angular momentum is an intrinsically quantum mechanical property that does not have a classical analog. All subatomic particles are spin ½ particles. The nucleus itself has a total spin angular momentum formed by the coupling of the individual spin angular momenta of its constituent protons and neutrons. The total nuclear spin angular momentum quantum number I may therefore take values: 0, 1/2, 1, 3/2, 3, 5/2, etc. A nucleus with non-zero quantum numbers I behaves as a small magnet or magnetic dipole with a magnetic moment ITLμITL. The magnetic moment is an intrinsic property of the nucleus, and it is associated with the angular momentum of the nucleus. As a vector, nuclear magnetic moment has two properties: magnitude and direction. The magnitude of µ is quantised and given by eqn (2.1):
μ = yh [square root of I (I + 1)] (2.1)
where h = h/2π is Plank's constant and y is the magnetogyric ratio, an inherent property of the nucleus. This parameter is unique for each nucleus. When I=0, then μ is zero, and the nucleus does not have magnetic properties. Nuclei with μ = 0, such as 12C,16O,32S cannot be studied by NMR. The smallest magnetic moments belong to protons and is called the nuclear magneton, μN; its value is calculated from μN = eh/2mpc(e is the electric charge, mp is the mass of proton, and c is the velocity of light) to 5.0505×10-27 J T-1. Under the influence of an external magnetic field of intensity (strength)B0 fixed along the z-axis of a static Cartesian coordinate system, the magnetic moment assumes discrete orientations (the nuclear Zeeman effect) governed by the magnetic quantum number mI. The allowed values of mI are -I, -I+1, -I+2, ..., I-1, I, giving rise to 2I+1 possible orientations. Each orientation defines an energy level or state, with energy:
E = - μzB0 = - mIyhB0 (2.2)
with mI=+½ or -½ for nuclei with I=½.μz(= mIyh) is the projection of the magnetic moment along the z-axis. The orientation of μz with respect to B0 defines the nuclear energy states. For I=½, μz parallel to B0 defines the energy state with the lower energy, whereas its anti-parallel orientation identifies the energy state with the higher energy. It is traditional to label the low and high-energy states with the Greek letters α and β, respectively. The α state with mI=+½ is often described as 'spin up', and the β state with mI=-½ as 'spin down'(Figure 2.1).
The energy space (eqn (2.3)) between the two states is:
[FORMULA NOT REPRODUCIBLE IN ASCII] (2.3)
where, Eβ and Eα are the energies of the upper and lower energy state, respectively. One important conclusion derived from eqn (2.3) is that the energy gap is variable and increases with increasing magnetic field strength B0(Figure 2.1). Transitions between energy states are induced by the magnetic component of the electromagnetic irradiation emitting in the radiofrequency (RF) region and having the same energy with the energy gap between the states. Excitation occurs provided that the NMR selection rule (ΔmI=±1) is satisfied. The radiation frequency inducing transitions for a given type of nucleus is given by eqn (2.4).
[FORMULA NOT REPRODUCIBLE IN ASCII] (2.4)
This equation indicates that at a given magnetic field strength each type of nucleus has its own resonance frequency, inasmuch the magnetogyric ratio is unique for each nucleus. Consequently, different types of nuclei cannot be detected in the same NMR experiment. Table 2.1 summarises the magnetic properties of selected nuclei.
Each food sample contains a huge number of molecules (1 mole contains 6.023×1023 molecules) and each molecule may comprise a large number of nuclei that occupy the available energy states. At room temperature, under the influence of the magnetic field B0, states of lower and higher energy are both occupied by nuclei (Figure 2.2).
This occurs because the energy difference between the two states is more than 100 times smaller than the energy of the thermal motion (kBT) of the molecules in solution at room temperature. The equilibrium population of the nuclear spins in the energy states is governed by the Boltzmann distribution law with a slight excess of nuclei in the state with the lower energy. According to Boltzmann law, the relative populations of the lower (Nα) and higher (Nβ) energy levels at room temperature (eqn (2.5)) are given by:
[FORMULA NOT REPRODUCIBLE IN ASCII] (2.5)
At T=303 K and magnetic field strength of 7.05 T, only 25 nuclei out of 1 million are in excess in the lower energy state. Nevertheless, this tiny excess of nuclei at the low energy state is the reason of the occurrence of the NMR phenomenon and at the same time the main cause of the low sensitivity characterising the NMR spectroscopy. The NMR sensitivity is lower compared to other branches of spectroscopy (e.g. UV-vis, IR, Raman). However, the low sensitivity of NMR is in part compensated by the wealth of information that can be extracted from typical NMR spectra. The sensitivity for nuclei with I=½ is measured by the signal-to-noise ratio, S/N, which depends, among others, on the magnetic field strength B0, the gyromagnetic ratio, the natural abundance of the nucleus, the temperature of the experiment, and the sample concentration (the population of nuclei at the lower state). According to eqn (2.5), the sensitivity is higher for nuclei with higher gyromagnetic ratio and natural abundance. For a given nucleus, the sensitivity is enhanced by the increasing magnetic field strength B0(eqn (2.5)) or by decreasing the temperature. The sensitivity is also augmented by manipulating other factors including the instrumental setup and data processing (see chapters 3 and 5).
In principle, all nuclear spins with magnetic properties (acquiring magnetic moment) can be studied by NMR. The most popular nuclei used in food analysis are 1H,13C, and to a lesser extent 15N, 19F, and 31P. The nuclear charge of these nuclei with I=½ is symmetric with a relatively low magnetic moment and affords signals with narrow line widths. This property facilitates greatly the interpretation of complex NMR spectra. The 1H nucleus, characterised by high gyromagnetic ratio and natural abundance (99.99%), is the most preferred nucleus used in food analysis. 1H NMR spectroscopy provides valuable information about minor components of foods (e.g. phenolic compounds, sterols, terpenes) in a reasonable experimental time. On the other hand, 1H NMR spectra of most foodstuffs are quite complex. Strong signal overlaps and scalar couplings within a relatively narrow range (15 ppm) of resonance frequencies make their interpretation a difficult task. Fortunately, the use of modern 2D NMR spectroscopy has largely mitigated this problem.
The13C nucleus is the second commonly used nucleus in the NMR analysis of foods.13C NMR spectroscopy provides additional information to that already obtained byH NMR, and can be considered as a complementary technique, which could facilitate the interpretation of complex1H NMR spectra. The much wider range (250 ppm) of13C resonance frequencies and the fact that13C NMR experiments are conducted with proton decoupling results in simplified spectra even for complex food matrices. Moreover, the phenomenon of NOE that is developed during proton decoupling increases the signal intensities of the protonated carbons by a factor of ca. 2 (see below). The main disadvantage of13C nucleus relative to1H nucleus is its lower sensitivity due to its lower magnetogyric ratio (by a factor of ca. 4) relative to proton, and its low natural abundance (ca. 1.1%). Consequently, recording a13C spectrum needs much more analytical time to show signals of minor components. This problem becomes even worse for quantitative analysis, where the long relaxation time of 13C nucleus should be taken into consideration in setting up the quantitative NMR experiment (see section 5.2).
Some of the problems showed by1H and13C NMR are removed by using the 31P nucleus, when possible. The large range of chemical shifts (ca. 1000 ppm) reported for the 31P nucleus ensures a good separation of signals obtained under proton decoupling, whereas its 100% natural abundance (the phosphorus atom has only one isotope), and its high relative sensitivity which is only ca. 15 times less than that of proton, make 31P NMR experiments a reliable analytical tool to determine amounts of the order of µmol, or lower, depending on the available instrumentation. Limitation of this method may be the fact that 31P nucleus is not widespread in foods as proton and carbon. As we shall see in section 7.1.2, a novel 31P NMR methodology developed recently has extended the application of this nucleus to components bearing no phosphorus.
(Continues...)
Excerpted from NMR Spectroscopy in Food Analysis by Apostolos Spyros, Photis Dais. Copyright © 2013 Apostolos Spyros and Photis Dais. Excerpted by permission of The Royal Society of Chemistry.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
