SEARCH

SEARCH BY CITATION

Keywords:

  • Crystallised honey;
  • near-infrared spectroscopy;
  • water activity;
  • water structure

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgment
  8. References

To characterise the changes occurring in honey under the influence of crystallisation process near-infrared spectroscopy has been used. The technique involved analysing the same type of honey by subtracting its crystallised spectrum from the liquid spectrum. The comparison of the difference spectra made it possible to determine specific water bonding processes affected by crystallisation. The surface area under the peak obtained during absorption by water was in the range of wavenumber ν ? 〈5330; 4965〉 cm−1 with the maximum at ν = 5155 cm−1. It was found out that the area under the peak was directly related to water activity changes that occurred in honey during the crystallisation process. A regression equation between the area under the peak and water activity assumed a linear form characterised by a high value of the determination coefficient R2.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgment
  8. References

When analysing the chemical composition of honey, we usually pay attention to the dominant content of the monosaccharide, i.e. glucose and fructose, which can amount up to 60–85% of the product mass (Lazaridou et al., 2004). Water is usually considered to be merely a supplement of the chemical composition of honey. However, the water content in honey ranges generally between 14 to do 20% and determines the basic properties of the product (Bakier, 2006). It should be noted here that the water content per se does not constitute the satisfactory parameter that determines its ‘nature’ in honey or other food products (Mathlouthi, 2001). A much more significant parameter, in fact, is the water activity, as it decides about honey’s storage and durability (Gleiter et al., 2006). The values of water activity in honey result mainly from the interactions between water particles and carbohydrates (Rüegg & Blanc, 1981). There are a number of reports describing the influence of common sugars on water activity (Rüegg & Blanc, 1981; Gaïda et al., 2006). It is well known that an increase of water content in liquid honey will cause a linear increase of its water activity (Chirife et al., 2006). Moreover, it should be noted that the interactions between honey carbohydrates and water will also affect changes of the water structure itself (Lewicki, 2004; Giangiacomo, 2006). These changes can be defined as being structural in nature, especially at the concentration higher than 30% of the sugar mass in the solution (Mathlouthi et al., 1980).

When stored, honey crystallises. The process is due to glucose that occurs in almost all types of honey in a supersaturated state (White, 1978; Assil et al., 1991; Cavia et al., 2002; Lazaridou et al., 2004). As a result of the crystallisation process, a two-phase mixture composed of glucose monohydrates and liquid plasma is formed in honey (White, 1978; Bhandari et al., 1999; Bakier, 2007a). The crystals may assume various shapes. Being mechanically pressed they form a crystalline suspension in which state they show pseudo plastic properties. Images of crystallised honey including its crystal shapes and surface areas (usually below 300 μm2) as well as the honey’s rheological properties are described by Mora-Escobedo et al. (2006). The thickness of honey crystals is of the order of several hundreds of nanometres (Bakier, 2003). The crystallised honey can be described as a colloidal suspension that is characterised by high viscosity of the order of about several hundreds of Pascal seconds. In the mixture, there exist strong surface interactions between the solid and liquid phases.

The liberation of the crystalline glucose causes a thinning of the remaining liquid phase and water bonding changes, which can be observed in an increase of water activity in honey (Gleiter et al., 2006; Zamora & Chirife, 2006). The increase of water activity has a far-reaching impact on the product durability, as exceeding the value of aw = 0.6 will enable the growth of osmophilic yeast and honey fermentation in its upper layers (Rüegg & Blanc, 1981; Sanz et al., 1994; Gleiter et al., 2006).

The physical manifestations of changes taking place in the water bonding have been usually analysed using NIR (near-infrared spectroscopy) spectra (Luck, 1974; Franks, 1983; Osborne & Fearn, 1986; Suzuki, 2002). This spectroscopy has a high accuracy of performed analyses, it is quick and there is no need for special sample preparation (Blanco & Villarroya, 2002). The method is also used to identify changes that occur not only in water structure under the influence of temperature (Czarnik-Matusewicz et al., 2005) but also in various chemical substances and carbohydrates (Berentsen et al., 1997;Frost & Molt, 1997; Li et al., 2003; Nørgaard et al., 2005; Giangiacomo, 2006). NIR spectroscopy is also applied to analyse the crystallisation process as well as crystalline substances (Buckton et al., 1998; Yu et al., 2004). The analysis methods used in spectroscopy can be classified into direct and indirect methods. The direct methods involve using rough spectroscopic spectra in order to identify the changes that occur in them under the influence of some specific factors (Luck, 1974; Osborne & Fearn, 1986). On the contrary, the indirect methods subject the spectra to various operations such as the calculation of the first and second derivative. (Hong et al., 1996; Zhou et al., 1998; Büning-Pfaue, 2003; Cozzolino et al., 2004). Alternative operations also include spectra subtractions and analysis of difference spectra obtained under the influence of various factors phenomena (Berentsen et al., 1997; Giangiacomo, 2006). Still other methods rely on the analysis of the global spectrum distribution into primary vibrations (Li et al., 2003; Czarnik-Matusewicz & Pilorz, 2006). It should be noted that the investigations concerning water content are concerned with some particular infrared spectra typical for radiation absorption by water: λ1 = 970 nm, λ2 = 1190 nm, λ3 = 1450 nm andλ4 = 1940 nm (Luck, 1974). Special attention must be paid to the spectrum range λ ? (1850; 2250), in which, apart from water absorption peak at λ5 = 2100, there is another peak that of radiation absorption by carbohydrates (Osborne & Fearn, 1986; Büning-Pfaue, 2003; Claude & Ubbink, 2006). This corresponds to a spectrum range within the wavenumber ν = (5585, 4500) cm−1, at which the temperature influence on the water structure is particularly well observed (Czarnik-Matusewicz & Pilorz, 2006).

The main purpose of this paper is to present the changes that occur in water bonding in honey as influenced by the crystallisation process. The investigation involves analysis of NIR spectra within the range of ν ? (5500; 4300) cm−1. The paper makes an attempt to combine the results obtained by using NIR spectroscopy with the data from water activity measurements.

Material and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgment
  8. References

The spectra of the investigated honeys were obtained using Nexus FT-IR spectrometer (Nicolet Instrument Corporation, Madison, WI, USA). The metre equipped with a helium–neon laser light source and Michelson’s interferometer made it possible to obtain continuous transmittance and absorbance spectra in the range of the wavenumber of ν ? 〈10000; 4000〉 cm−1. For this purpose, Fourier transformation was used. The equipment included an InGaAs detector and XT-KBr beamsplitter. The spectroscope was connected to Pentium II PC and controlled by a dedicated software packet omnic E.S.P 5.1 (NIC, Madison, WI, USA). All measurements were performed using a 0.1 mm path-length quartz cuvette. A total of fifty scans were averaged for each spectrum.

Water activity was measured at a constant temperature of 25 °C controlled by AQUA LAB CX 2 device (Aqua Lab Industries, Garden Grove, CA, USA). The measurements were conducted at the accuracy of Δaw = 0.003 and repeated three times for each sample. The content of water in the honey was determined using the refractometric method using Abbe refractometer (Carl Zeiss, Jena, Germany) measuring the light refraction coefficient.

In the investigations, ten different types of crystallised honeys were used. Each sample of honey was divided into two parts. One part was investigated in the crystallised form whereas the other was subjected to liquefaction by heating the hermetically sealed samples to 55 °C for 24 h. The samples, having been being cooled to room temperature, were subjected to spectroscopic measurements and also water activity analysis at 25 ± 1 °C.

The spectroscopic investigations involved obtaining the spectra of the same sample type both in liquid and crystalline states. In the research, a single drop of liquid or crystallised honey was placed on a quartz cuvette and then another identical cuvette was pressed against it. Prior to that, 0.1 mm separator was inserted between the cuvettes. The layer thickness of 0.1 mm was selected after a series of tests. An increase of the thickness of the layer caused a considerable increase of the absorbance value. The crystallised honey used for the investigations was the mixture of two phases and no attempt was made to separate the phases. The samples of crystallised honey used for investigations were obtained by mechanical crashing and mixing up the original honey structure after crystallisation. A very important element of the research was the application of the identical mass of the crystallised honey drops placed on the cuvette. The spectra obtained for several scores of samples of the same type of honey were compared. To create the difference spectra, two characteristic spectra were chosen from among the crystallised and liquid honey samples. The spectra usually overlapped.

Next, the spectra of crystallised samples were subtracted from the liquid state sample spectra of the same types of honey. The resulting difference spectra showed some characteristic changes that occurred in the samples during the crystallisation process. The surface area of the peak was then measured in the range of ν ? 〈5330; 4965〉 cm−1. Finally, the surface area obtained was compared with the increase of water activity past crystallisation.

Results and discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgment
  8. References

Table 1 presents the parameters of the investigated types of honey. Their water content ranges from 17.3% to 19.8% and water activity values vary from 0.556 to 0.596 in the liquid state. The results confirm the general principle stating that water activity in honey is lower than 0.6 at the water content below 20%. Unfortunately, the crystallisation process increases water activity. The samples of the same types of crystallised honey are characterised by water activity between 0.585 and 0.639. An average increase of water activity caused by crystallisation amounts to Δaw = 0.038 and is in good agreement with the results reported by others (Gleiter et al., 2006; Zamora & Chirife, 2006).

Table 1.   Properties of the investigated honeys
Sample numberType of honeyLight refraction coefficientWater content %Water activity of liquefied honeyWater activity in crystallised honey
 1Fine-grained rape honey1.488019.40.5860.639
 2Fine-grained dandelion honey1.491817.90.5470.585
 3Fine-grained linden honey1.490818.850.5780.611
 4Coarse-grained linden honey1.487019.80.5960.637
 5Fir tree honeydew1.493217.30.5680.587
 6Coarse-grained multifloral honey1.487819.50.5860.633
 7Coarse-grained buckwheat honey1.491218.10.5560.59
 8Pine honey1.488219.80.5960.624
 9Fine-grained buckwheat honey1.489718.70.5670.609
10Fine-grained multifloral honey1.488619.250.5740.622

An increase of water content up to w = 19.5% gives an increase of water activity analogical to the values obtained by Gleiter et al. (2006) on a considerably larger number of honey samples. However, the samples characterised by water content close to 20% show a considerably lower increase of water activity and the results are analogical to the ones published by Rüegg & Blanc (1981).

Figure 1 shows the spectra of two different honey types: rape honey (sample no. 1) and buckwheat honey (sample no. 2) in liquid states and crystallised states. In the range of ν ? 〈5500; 4500〉, the spectra have two characteristic peaks at ν1 = 5155 cm−11 = 1940 nm) and ν2 = 4762 cm−12 = 2100 nm). The former peak shows the radiation absorption by water, whereas the latter shows the radiation absorption by carbohydrates (Osborne & Fearn, 1986; Rambla et al., 1997; Claude & Ubbink, 2006). Between the extremes, there is a minimum at ν3 = 4965 cm−1 for all spectra. The rape honey in the liquid state shows a higher water content (w = 19.4%) and when compared with the buckwheat honey (w = 18.7%) shows a considerably low absorbance value, i.e. ΔA = 0.093 at ν1 = 5155 cm−1.

The spectra of crystallised honeys are characterised by a higher absorbance value throughout the whole range of the wavenumber. This is due to a stronger radiation absorption by the crystals of monohydrate glucose (Bakier, 2007b). And in spite of the fact that they have analogical maximum values at ν1 = 5155 cm−1 and ν2 = 4762 cm−1 as well as minimum values at ν3 = 4965 cm−1 as the liquid honeys, but the differences between them are variable. To accurately illustrate the differences between the liquid and crystallised honeys, an operation of subtracting the crystallised honey spectrum from the liquid honey spectrum was performed. The results of operation for the honey samples presented on Fig. 1 are shown in Fig. 2.

image

Figure 1.  Spectra of both liquid and crystallised honeys: rape honey (Table 1, sample no. 1), buckwheat honey (Table 1, sample no. 9).

Download figure to PowerPoint

image

Figure 2.  Differential spectra obtained by subtracting the crystallised honey spectra from the liquid honey spectra (Table 1, sample nos 1 and 9).

Download figure to PowerPoint

Because of the fact that the crystallised honey spectra show higher absorbance values, the resulting difference spectra have negative absorbance value. A characteristic effect is the peak still observed at ν1 = 5155, i.e. on the wavenumber typical for the radiation absorption by water (Luck, 1974; Czarnik-Matusewicz & Pilorz, 2006). This region NIR corresponding to the water absorption band represent the changes in the water structure (Büning-Pfaue, 2003; Giangiacomo, 2006). The crystallisation process causes changes in structure and water bonding in honey. Water, apart from interacting in the solution that becomes less abundant in glucose after the crystallisation process, is also involved in surface interactions with crystals. The research work carried out by the author showed that surface interactions between water and honey crystals are significant and should be taken into consideration when studying changes in water activity in honey (Bakier, 2007a). Because of the above, the peak that is formed in the difference spectra of both liquid and crystallised honey is related to the change of water bonding processes in the product. By calculating the surface area under the peak, it is possible to determine the quantitative changes that occur in honey water bonding. As it was already stated in the introduction, crystallisation increases water activity in honey. Combining both of these phenomena, i.e. an increase of water activity and the surface area under the peak occurring in the difference spectra at the maximum of ν1 = 5155 cm−1, seems to be completely justified. Figure 3 presents the sections of difference spectra of the investigated honeys in the range of ν ? 〈5400; 4900〉 cm−1. All the difference spectra presented in Fig. 3 have local minimum values of absorbance at ν2 = 4965 cm−1. This value of the wavenumber was used as a limit value to calculate the surface area under the peak with the maximum at ν1 = 5155 cm−1. On the other side, ν3 = 5330 cm−1 was accepted as the interval limit value. The straight line running across these two points A and B (Fig. 3; sample no. 3) cuts through the spectrum at ν2 = 4965 cm−1 and is tangent at the point ν3 = 5330 cm−1. The surface area SA between the peak and the straight line AB determines the changes that occur in water under the influence of the honey crystallisation process. To calculate the SA value, omnic software was used. The physical interpretation of the quantity has made it possible to state that SA is the product of the absorbance and the wavenumber, which means that its unit is cm−1.

image

Figure 3.  The differential spectra in the range of wavenumber ν ? 〈5400; 4900〉 cm−1.

Download figure to PowerPoint

Figure 4 presents the chart showing the dependence of the surface area under the peak in the difference spectrum in the range of ν ? 〈5330; 4965〉 cm−1 with the maximum ν1 = 5155 cm−1 on the water activity increase in the honey samples after crystallisation. The result shown in Fig. 4 explicitly confirms a close dependence between water activity increase and the surface area under the peak in the spectrum within the range of ν ? 〈5330; 4965〉. The relation has a linear characteristics and a high coefficient of determinacy R2 = 0.9406 at the level of statistical significance of P ≤ 0.05. Additionally, it should be noted that initially we have a coordinate system which means that at zero value of water increase the surface area under the peak is equal to zero as there is no crystallisation. A factor interfering with the measurements involved the differences in the morphology of the crystalline structure of particular types of honey. Hence, the data concerning the crystallised honeys shown in Table 1 include the information on the crystalline structure of the samples. Literature reports on NIR spectroscopy have shown that this factor may cause some interference by changing the reflectance of the light beam hitting the samples. (Osborne & Fearn, 1986). However, in the investigations conducted here no considerable influence of reflectance on the results of the analyses was observed.

image

Figure 4.  Dependence between water activity increase and the surface area under the peak in the range ν ? 〈5330; 4965〉 cm−1.

Download figure to PowerPoint

Conclusions

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgment
  8. References

Recapitulating the results of the investigations, it should be noted, first of all, that the applied spectroscopic analysis made it possible to identify the quantitative relation between the increase of water activity and the changes affected by the crystallisation process. Using the subtraction technique of the crystallised honey spectrum from the liquid one, it was possible to identify the peak value of water absorption with its maximum at ν1 = 5155 cm−1. The surface area under the peak in the range of ν ? 〈5330; 4965〉 cm−1 correlated linearly with the increase of water activity caused by the crystallisation process and could be expressed by the relation: A = 315.61.Δaw with the determination coefficient R2 = 0.9406. The existence of the peak is due to the phenomenon of stronger water bonding in the liquid honey in contrast to the crystallised state.

Although water activity measurements in relation to water bonding changes have a global character, it should be observed, however, that the spectroscopic analysis of honey spectra enables us to obtain a considerable amount of additional data on the processes that occur in honey during crystallisation.

Analysing the shape of the peak in the difference spectrum with the maximum at ν1 = 5155 cm−1, it is possible to identify the changes that occur in the water structure caused by honey crystallisation. This can be performed by creating specific spectra originated by particular vibration types of OH groups, i.e. both stretching symmetric and asymmetric or bending (Luck, 1974; Czarnik-Matusewicz & Pilorz, 2006). Additional identification of the carbohydrate peak at ν1 = 4716 cm−1 makes it possible to analyse the changes occurring in the solution of carbohydrate responsible for forming the liquid phase in the crystallised honey. The data contained in the NIR spectra concerning the carbohydrate changes occurring in honey as a result of the crystallisation process are under investigation.

Acknowledgment

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgment
  8. References

The author wishes to express his gratitude to Professor Peter P. Lewicki for his support and encouragement in his research work using NIR spectroscopy. The research work was financed from the science budget for the years 2007–2008 – Polish Project Number N312 011 32/0755.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgment
  8. References
  • Assil, H.I., Sterling, R. & Sporns, P. (1991). Crystal control in processed liquid honey. Journal of Food Science, 56, 10341037.
  • Bakier, S. (2003). Grain size characteristics of the solid phase occurring at honey granulation. Inżynieria Rolnicza, 7, 510 (in Polish).
  • Bakier, S. (2006). Characteristics of water state in some chosen types of honey found in Poland. Acta Agrophysica, 7, 715.
  • Bakier, S. (2007a). Influence of glucose changes on water activity in selected honeys. Acta Agrophysica, 9, 719.
  • Bakier, S. (2007b). Investigation of interaction between glucose and fructose in aqueous mixtures. Polish Journal of Food and Nutrition Science, 57, 716.
  • Berentsen, S., Stolz, T. & Molt, K. (1997). Analysis of aqueous solutions by near-infrared spectroscopy (NIRS) IV. One- and two-component systems of organic compounds in water. Journal of Molecular Structure, 410–411, 581585.
  • Bhandari, B., D'Arcy, B. & Chow, S. (1999). Rheology of selected Australian honeys. Journal of Food Engineering, 41, 6568.
  • Blanco, M. & Villarroya, I. (2002). NIR spectroscopy: a rapid-response analytical tool. Trends in Analytical Chemistry, 21, 240250.
  • Buckton, G., Yonemochi, E., Hammond, J. & Moffat, A. (1998). The use of near infra-red spectroscopy to detect changes in the form of amorphous and crystalline lactose. International Journal of Pharmaceutics, 168, 231241.
  • Büning-Pfaue, H. (2003). Analysis of water in food by near infrared spectroscopy. Food Chemistry, 82, 107115.
  • Cavia, M.M., Fernandez-Muino, M.A., Gomez-Alonso, E., Montes-Perez, M.J., Huidobro, J.F. & Sancho, M.T. (2002). Evolution of fructose and glucose in honey over one year influence of induced granulation. Food Chemistry, 78, 157161.
  • Chirife, J., Zamora, M.C. & Motto, A. (2006). The correlation between water activity and % moisture in honey: fundamental aspects and application to Argentine honeys. Journal of Food Engineering, 72, 287292.
  • Claude, J. & Ubbink, J. (2006). Thermal degradation of carbohydrate polymers in amorphous states: a physical study including colorimetry. Food Chemistry, 96, 402410.
  • Cozzolino, D., Kwiatkowski, M.J., Parker, M. et al. (2004). Prediction of phenolic compounds in red wine fermentations. Analytica Chimica Acta, 513, 7380.
  • Czarnik-Matusewicz, B. & Pilorz, S. (2006). Study of the temperature-dependent near-infrared spectra of water by two-dimensional correlation spectroscopy and principal components analysis. Vibrational Spectroscopy, 40, 235245.
  • Czarnik-Matusewicz, B., Pilorz, S. & Hawranek, J.P. (2005). Temperature-dependent water structural transitions examined by near-IR and mid-IR spectra analyzed by multivariate curve resolution and two-dimensional correlation spectroscopy. Analytica Chimica Acta, 544, 1525.
  • Franks, F. (1983). Water. Pp. 2935. Cambridge: The Royal Society of Chemistry.
  • Frost, V.J. & Molt, K. (1997). Analysis of aqueous solutions by near-infrared spectrometry (NIRS) III. Binary mixtures of inorganic salts in water. Journal of Molecular Structure, 410–411, 573579.
  • Gaïda, L.B., Dussap, C.G. & Gros, J.B. (2006). Variable hydration of small carbohydrates for predicting equilibrium properties in diluted and concentrated solutions. Food Chemistry, 96, 387401.
  • Giangiacomo, R. (2006). Study of the water-sugar interactions at increasing suger concentration by NIR spectroscopy. Food Chemistry, 96, 371379.
  • Gleiter, R.A., Horn, H. & Isengard, H.D. (2006). Influence of type and state of crystallisation on the water activity of honey. Food Chemistry, 96, 441445.
  • Hong, J., Yamaoka-Koseki, S., Tsujii, Y. & Yasumoto, K. (1996). Applicability of near-infrared spectroscopic method to unfreezable water measurements in egg white lysozyme and soluble starch. Lebensmittel-Wissenschaft und-Technologie, 30, 406410.
  • Lazaridou, A., Costa, G.B., Bacandritsos, N. & Sabatini, A.G. (2004). Composition, thermal and rheological behaviour of selected Greek Honeys. Journal of Food Engineering, 64, 921.
  • Lewicki, P.P. (2004). Water as the determinant of food engineering properties. A review. Journal of Food Engineering, 61, 483495.
  • Li, R., Jiang, Z., Shi, S. & Yang, H. (2003). Raman spectra and 17O NMR study effects of CaCl2 and MgCl2 on water structure. Journal of Molecular Structure, 645, 6975.
  • Luck, W.A.P. (1974). Structure of Water and Aqueous Solutions. Pp. 220284. Weinheim: Verlag Chemie.
  • Mathlouthi, M. (2001). Water content, water activity, water structure and the stability of foodstuffs. Food Control, 12, 409417.
  • Mathlouthi, M., Luu, C., Meffroy-Biget, A.M. & Luu, D.V. (1980). Laser-Raman study of solute–solvent interactions in aqueous solutions of D-fructose, D-glucose, and sucrose. Carbohydrate Research, 81, 213223.
  • Mora-Escobedo, R., Moguel-Ordóňez, Y., Jarmillo-Flores, M.E. & Gutiérrez-López, G.F. (2006). The composition, rheological and thermal properties of tajonal (viguiera denata) Mexican honey. International Journal of Food Properties, 9, 299316.
  • Nørgaard, L., Hahn, M.T., Knudsen, L.B., Farhat, I.A. & Engelsen, S.B. (2005). Multivariate near-infrared and Raman spectroscopic quantifications of the crystallinity of lactose in whey permeate powder. International Dairy Journal, 15, 12611270.
  • Osborne, B.G. & Fearn, T. (1986). Near Infrared Spectroscopy in Food Analysis. Pp. 3742. Avon: Longman Scientific and Technical.
  • Rambla, F.J., Garrigucs, S. & De La Guardia, M. (1997). PLS-NIR determination of total sugar, glucose, fructose, and sucrose in aqueous solutions of fruit juices. Analytica Chimica Acta, 344, 4153.
  • Rüegg, M. & Blanc, B. (1981). The water activity of honey and related sugar solutions. Lebensmittel-Wissenschaft und-Technologie, 14, 16.
  • Sanz, S., Gradillas, G., Jiemeno, F., Perez, C. & Juan, T. (1994). Fermentation problem in Spanish north-coast honey. Journal of Food Protection, 58, 515518.
  • Suzuki, T. (2002). Hydrogen-bond studies of thin film water using near-infrared spectroscopy in the 970 nm spectral region. Applied Surface Science, 187, 261265.
  • White, J.W. Jr (1978). Honey. Advances in Food Research, 24, 288374.
  • Yu, L.X., Lionberger, R.A., Raw, A.S., D’Costa, R., Wu, H. & Hussain, A.S. (2004). Applications of process analytical technology to crystallization processes. Advanced Drug Delivery Reviews, 56, 349369.
  • Zamora, M.C. & Chirife, J. (2006). Determination of water activity change due to crystallization in honeys from Argentina. Food Control, 17, 5964.
  • Zhou, X., Hines, P. & Borer, M.W. (1998). Moisture determination in hygroscopic drug substances by near infrared spectroscopy. Journal of Pharmaceutical and Biomedical Analysis, 17, 219225.