Caramelization Processes in Sugar Glasses and Sugar Polycrystals

Home , Caramelization, Sugar glass, Trehalose

New Physics: Sae Mulli (The Korean Physical Society), DOI: 10.3938/NPSM.62.761 Volume 62, Number 7, 2012¸ 7Z4, pp. 761∼767

Jeong-Ah Seo · Hyun-Joung Kwon · Dong-Myeong Shin · Hyung Kook Kim · Yoon-Hwae Hwang∗ Department of Nanomaterials Engineering & BK21 Nano Fusion Technology Division, Pusan National University, Miryang 627-706

(Received 3 April 2012 : revised 3 June 2012 : accepted 2 July 2012)

We studied the chemical dehydration processes due to the caramelization in sugar glasses and sugar polycrystals. The dehydration processes of three monosaccharide sugars (fructose, galactose, and glucose) and three disaccharide sugars (sucrose, maltose, and trehalose) were compared by using a thermogravimetic-diﬀerential thermal analyzer to measure the mass reduction. The amounts of mass reductions in sugar glasses were larger than those in sugar polycrystals. However, the amount of mass reduction in trehalose glasses was smaller than that in trehalose polycrystals. This unique dehydration property of trehalose glasses may be related to the high glass transition temperature, which might be related to a superior bioprotection ability of trehalose.

PACS numbers: 87.14.Df, 81.70.Pg, 64.70.P Keywords: Dehydration, Caramelization, Sugar glass, Trehalose, Disaccharide

I. INTRODUCTION higher than that of other disaccharides [6,7]. Therefore, the viscosity of trehalose is higher than that of other There are many strategies in nature for the long-term sugars at a given temperature. Several researchers have survival and storage of organisms, and a bio-protection pointed out that the high glass-transition temperature of effect is one of the most interesting among those long- trehalose may contribute to the preservation of biologi- term survival and storage mechanisms. The stabilizing cal molecules through the control of water mobility [8]. role of sugars on dehydrated biological molecules has The flexibility of α, α-(1→1)-glycosidic linkage in tre- been recognized for many years in biological, pharma- halose could also be an important clue in explaining its ceutical, and food sciences [1–5]. Sugar is a main con- biological functions [9–11]. And trehalose has a hydra- stituent of the biological system and the systems of sugar tion characteristic which displays extremely interesting and sugar containing materials are a matter of common features, including the ability to protect and reversibly interest to many researchers. Sugars also have a great reconstitute proteins and bio-membranes from dehydra- significance in nature as a glass-forming material because tion and freeze-drying. There are many studies about the nature makes use of glasses to preserve biological tis- protection ability of trehalose during drying and ensuing sues in the dehydrated state. The high viscosity of sugar storage [12–14]. These studies reported several possible glasses leads to an increased stability of preserved mate- origins of their protection ability, which include water rials. replacement processes [15–17], vitrification [18,19], and Alpha, alpha - trehalose (α-D-glucopyranosyl α-D- dynamic reducers [11]. glucopyranoside) is a well-known, non-reducing disac- Sugar is easily caramelized by heating and this chemi- charide that is commonly found in yeast, fungi, bacte- cal degrade is always occurred with dehydration. In this ria, mushrooms, and desert plants. Trehalose consists of study, we measured the chemical dehydration which was two glucose molecules and has several unique character- istics. The glass transition temperature of trehalose is induced by caramelization process in the glassy and polycrystalline phase of trehalose and several other sugars. ∗E-mail: [email protected] This is important to understand not only the protection -761- -762- ôÇ²DGÓütoÆpolysaccharides. We compared the dehydration processes of sugar glasses and sugar crystals by measuring the chemical dehydration during the heating and found a unique dehydration process in trehalose. Fig. 1. The X-ray diffraction spectra of trehalose dihydrate vs. drying time (Th: Trehalose dihydrate, Tγ : II. EXPERIMENTS Partially hydrated trehalose, Tβ: Anhydrous trehalose).

In this study, we used three monosaccharides (glucose, of the sample cell was 5.2 mm in diameter and 5.1 mm in galactose, and fructose) and three disaccharides (sucrose, height. Al2O3 was used as a reference material. The ini- trehalose dihydrate, and maltose monohydrate). The tial weight of the samples was 15 mg. All measurements trehalose dihydrate was donated by the Cargill Corpo- were repeated more than 5 times. ration. All other sugars (glucose, galactose, fructose, su- TG-DTA is an effective tool to heat the sample uni- crose, and maltose monohydrate) were purchased from formly and quickly [20,21]. Glass phases of dried sugars the Sigma Chemical Company and were used without were prepared using a TG-DTA because sugars are ex- further purification. Table 1 shows the chemical struc- tremely sensitive to heat and can easily be changed into tures, molecular weights, and melting temperatures of caramel due to the dehydration process [22]. The glass sugars. Trehalose and maltose exist in hydrate forms at transition temperatures of sugar glasses were measured room temperature and sugar has a deliquescence charac- using a differential scanning calorimeter (DSC; MAC sci- teristic. Therefore, we dried all sugar samples very care- ence, DSC3100, Japan) and the glass phases of sugars fully before the experiments because the caramelization were also confirmed using an X-ray powder diffraction process could be affected by the moisture contained in (XRD) method. The DSC data were measured during the sugars. First, we spread the sugars on an aluminum heating with four different heating rates of 2, 4, 6, and 8 pan with balance which has ± 0.001 g precision and ◦C/min. The glass transition temperatures were deter- ◦ heated for 2 hours at 110 C by using ceramic IR heat- mined as the midpoint between the onset and the end ing elements to prepare anhydrous samples. The states of point in the heat flow vs. the temperature plot. all sugar samples were confirmed using the X-ray powder diffraction method (Rigaku GDX-1193A, JAPAN) with Cu-K radiation. We dried the sugar samples again using III. RESULTS AND DISCUSSION a thermo gravimetric-differential thermal analyzer (TG- DTA) more than 3 hours before the dehydration mea- Trehalose exists in dihydrate form at room tempera- surements until mass of the samples were stabilized and ture and can be transformed into either a partially hy- the mass reductions were measured immediately to pre- drated Tγ phase or an anhydrous crystalline Tβ phase. vent the rehydration. During the whole procedure of the As we mentioned in the experimental part, we dried the second drying and measurement in TG-DTA, the sample sugar samples very carefully before measurements. Es- was under the nitrogen atmosphere. All sugar samples pecially, trehalose was known as the most intimate ma- were heated at a constant heating rate of 2 ◦C/min. A terial with water [15–17]. We confirmed the anhydrous cylindrical-shaped aluminum cell was used and the size crystal phases of trehalose by using an XRD analysis Caramelization Processes in Sugar Glasses and Sugar Polycrystals – Jeong-Ah Seo 1px -763-

Table 1. The chemical formulas, molecular weights and melting temperatures of sugars.

Chemical Molecular Weight Melting temperature Sugar Formula (g/mol) (◦C) Fructose 119 ∼ 122

Monosaccharide Glucose C6H12O6 180.2 153 ∼ 156 Galactose 170 Sucrose 185 ∼ 187

Disaccharide Maltose C12H22O11 342.3 120 ∼ 140 Trehalose 213

Fig. 2. The TG-DTA data of trehalose. The solid, dashed, and dotted lines represent the mass reduction of Fig. 3. The X-ray diﬀraction patterns of trehalose glass trehlaose, the DTA curve of trehalose dihydrate Tγ , and and anhydrous polycrystal (Tβ). the DTA curve of anhydrous trehalose Tβ, respectively.

because the weight fraction of the hydrated water in tre- method [23–25]. Fig. 1 shows the change of the XRD halose dihydrate is about 10 wt.%. The dotted line is patterns of polycrystalline trehalose at a different drying the DTA curve of the T phase of trehalose. We found time. At the 5 minutes heating, the X-ray diffraction h two endothermic peaks at 100 ◦C and 118 ◦C. These two pattern of the trehalose shows dihydrate Th phase and it peaks correspond to the evaporation of the water in the means phase of trehalose did not change during the dry- trehalose dihydrate. The peak around 190 ∼ 200 ◦C rep- ing. After drying for 10 minutes, the trehalose dihydrate resents the melting temperature of trehalose. The dashed T phase changed into a partially hydrated T phase. It h γ line shows the DTA curve of anhydrous trehalose. In the finally changed into an anhydrous crystalline Tβ phase Tβ phase, we could not observe an evaporation peak at after 20 minute heating. temperatures ranging from 100 to 120 ◦C and found only To analyze the drying process of trehalose more pre- a melting peak at temperatures around 211 ◦C indicat- cisely, we measured the drying process of trehalose dihy- ing the melting temperature was affected by the contents drate by using the TG-DTA and the results are shown of moisture. in Fig. 2. The solid line represents the mass reduc- The glass phases of sugars were also confirmed by us- tion of trehalose dihydrate with increasing temperature. ing a XRD method. Fig. 3 shows the XRD patterns of The mass of trehalose started to decrease at temperature glass and polycrystalline phases of anhydrous trehalose ◦ around 80 C and reduced to about 1.5 mg at temper- Tβ. The diffraction data of polycrystalline trehalose Tβ ature around 125 ◦C. The initial mass of trehalose di- showed its own characteristic pattern and these data hydrate was 15 mg and the mass of trehalose dihydrate were consistent with the data in references [23–25]. On was reduced to about 10 wt.%. This is a reasonable value the other hand, there was no distinct structure peak in -764- ôÇ²DGÓütoÆ

Fig. 4. The DSC data of sugars with four different heating rates. (a) Fructose, (b) Glucose, (c) Galactose, (d) Sucrose, (e) Maltose, and (f) Trehalose. The glass transition temperatures were taken to be the mid-point between the onset and end temperatures. the diffraction data of trehalose glass. The broad first temperatures. The glass transition temperature of fruc- peak at around 2θ ≈ 20◦. represents the average near- tose, glucose, galactose, sucrose, maltose, and trehalose est neighbor distance between trehalose molecules. The are 7.8 ∼ 8.6 ◦C, 32.7 ∼ 36.0 ◦C, 34.5 ∼ 37.8 ◦C, 68.9 ∼ glass phases of other sugars were also confirmed by using 71.9 ◦C, 72.6 ∼ 75.9 ◦C, and 107.1 ∼ 110.3 ◦C, respec- XRD method [20]. tively. The glass transition temperature of trehalose was The glass phases of sugars were also confirmed by mea- the highest among the sugars. Therefore, the viscosity suring the glass transition temperatures by using a dif- of trehalose is highest among the sugars at a given tem- ferential scanning calorimeter (DSC). Figs. 4(a) ∼ 4(f) perature. The high viscosity of glass and its concomitant show the glass transition temperatures of fructose, glu- low-molecular mobility are known to preserve the stabil- cose, galactose, sucrose, maltose, and trehalose, respec- ity of protein [26]. Green and Angell have concluded that tively. The heating rates used in these measurements the bio-protection from destruction by freezing is related ◦ were 2, 4, 6, and 8 C/min. The glass transition tem- to the high Tg of trehalose [6,7]. perature increased with increasing heating rate due to We measured the mass reductions of glassy and poly- a thermal delay effect. The glass transition temperature crystalline sugars using a TG-DTA with a constant heat- was taken to be the mid-point between the onset and end ing rate of 2 ◦C/min in the temperature range which Caramelization Processes in Sugar Glasses and Sugar Polycrystals – Jeong-Ah Seo 1px -765-

Fig. 5. The mass reductions of sugar glasses and polycrystals during the heating. (a) glucose, (b) fructose, (c) galactose, (d) sucrose, (e) maltose, and (f) trehalose. The straight lines represent sugar glasses and the dotted lines represent sugar polycrystals. The vertical straight lines represent the melting temperatures of each sugar. covers the glass transition temperature and the melting glasses. And sugar glasses showed durable mass reduc- temperature of each sugar. Figs. 5(a) ∼ 5(f) respec- tion from temperatures lower than melting temperatures tively show the mass changes of glucose, fructose, galac- while sugar crystals showed rapid mass reduction at tem- tose, sucrose, maltose, and trehalose during the heating. peratures around the melting temperatures. The straight lines and dotted lines represent the mass Diﬀerent from other sugars, the mass reduction of tre- of sugar glasses and sugar polycrystals. The vertical halose shows a diﬀerent trend as shown in Fig. 5(f). The straight lines represent the known melting temperatures mass of trehalose crystals started to decrease at temper- of each sugar. In all sugar glasses and sugar polycrystals, atures around 140 ◦C and the trends of curve changed at the mass abruptly reduced at temperatures around the temperatures around 180 ◦C and 220 ◦C. As mentioned melting temperature due to the dehydration which was in the experiments, all samples were dried very carefully induced by a caramelization process. We can observe in and the evaporation of water in trehalose dihydrate must Figs. 5(a) ∼ 5(e) that the amount of mass reductions start at temperatures around 100 ◦C as shown in Fig. 2. in sugar polycrystals was smaller than those in sugar Moreover, the dehydration processes of maltose glass and -766- ôÇ²DGÓütoÆ

Fig. 6. Heat flow (solid line) and mass (dotted line) of trehalose polycrystal. polycrystal in Fig. 5(e) did not show any step and was the same to that of other sugars though they also exist in hydrated form at room temperature like a trehalose. Therefore, the observed three-step behavior in trehalose polycrystal is not due to the evaporation of remaining water in trehalose. Fig. 6 shows the heat flow (solid line) and mass (dotted line) of trehalose polycrystal together. In the heat flow data of trehalose polycrystal, two pre- ◦ melting endothermic peaks around temperatures 170 C Fig. 7. The mass of (a) sugar polycrystals and (b) sugar ◦ and 190 ◦C were found and steps in mass data coincide glasses at the temperatures ranging from 80 C to 240 ◦C. with those two pre-melting endothermic peaks. A more significant difference in the dehydration process of trehalose compared with other sugars was that the amount glass transition could be related to the bio-protection of mass reduction of trehalose glass was smaller than that from disturbance of freezing. of the trehalose crystal. The weight percent of the mass reductions of sugar polycrystals and glasses were compared in Figs. 7(a) IV. CONCLUSIONS and 7(b). It is interesting that the mass reduction of tre- We studied the chemical dehydration processes which halose glass was the lowest among all other sugar glasses were induced by caramelization in glassy and poly- at whole temperatures ranging from 80 ◦C to 240 ◦C while trehalose crystal showed relatively high mass re- crystalline sugars. The dehydration processes of three duction. The lowest mass reduction of trehalose glass in- monosaccharide sugars (fructose, galactose, and glucose) dicates that the trehalose glass is chemically stable com- and three disaccharide sugars (sucrose, maltose, and tre- pared to the other sugar glasses under the heating or dry- halose) were compared by measuring the mass reducing. These results are consistent with the non-reducing tion by using a thermogravimetic-differential thermal an- character and low tendency of hydrolyzation properties alyzer (TG-DTA) in the temperature range which cov- of trehalose. It also might be related to the high glass ers the glass transition temperature and melting tem- transition temperature of trehalose. In 1992, Karmas perature of each sugar. We found that in all sugars ex- et al. found that the non-enzymatic browning reactions cept trehalose, the mass reductions of sugar polycrystals (caramelization) in food systems were slowed down by are smaller than that of sugar glasses. However, in the the high glass transition temperature [27]. And the high case of trehalose, the mass reduction in the glass phase Caramelization Processes in Sugar Glasses and Sugar Polycrystals – Jeong-Ah Seo 1px -767- was smaller than that in the crystalline phase, indicat- [10] M. Kuttel and K. J. Naidoo, Carbohydr. Res. 340, ing that the dehydration process in the trehalose glass 875 (2005). was suppressed and chemically stable compared to the [11] Y. Choi, Y, K. W. Cho, K. Jeong and S. Jung, Car- other sugar glasses under the heating or drying. The ob- bohydr. Res. 341, 1020 (2006). served unique dehydration process in trehalose glasses is [12] J. H. Crowe, L. M. Crowe and D. Chapman, Science related to the high glass transition temperature which is 223, 701 (1984). strongly related to a superior bioprotection ability. [13] K. Koster, M. S. Webb, G. Bryant and D. V. Lynch, Biochim. Biophys. Acta. 1193, 143 (1994). ACKNOWLEDGEMENTS [14] M. F. Mazzobre, M. P., Buera and J. Chirife, Biotechnol. Prog. 13, 195 (1997). We thank the late H. Z. Cummins for suggesting sug- [15] P. Bordat, A. Lerbret, J.-P. Demaret, F. Affouard ars for a glass transition study. We also thank to Su Jae and M. Descamps, Europhys. Lett. 65, 41 (2004). Kim, Hyun-Woo Choi, and Yong-Suk Yang for support- [16] A. Lerbret, P. Bordat, F. Affouard, Y. Guinet, A. ing the TG-DTA and DSC measurements. This work Hédoux, L. Paccou, D. Prévost and M. Descamps, was supported by a 2-Year Research Grant of Pusan Na- Carbohydr. Res. 340, 881 (2005). tional University. [17] R. Giangiacomo, Food Chemistry 96, 371 (2006). [18] J. H. Crowe, S. B. Leslie and L. M. Crowe, Cryobi- REFERENCES ology 31, 355 (1994). [19] J. H. Crowe, J. F. Carpenter and L. M. Crowe, [1] J. Carpenter, L. M. Crowe and J. H. Crowe, Annu. Rev. Physiol. 60, 73 (1998). Biochim. Biophys. Acta. 923, 109 (1987). [20] J.-A. Seo, J. Oh, D. J. Kim, H. K. Kim and Y.-H. [2] L. M. Crowe, J. H. Crowe, J. Carpenter and C. A. Hwang, J. Non-Cryst. Solids 333, 111 (2004). Wistrom, Biochem. J. 242, 1 (1987). [21] J.-A. Seo, S. J. Kim, J. Oh, Y. S. Yang, H. K. Kim [3] S. Rossi, M. P. Buera, S. Moreno and J. Chirife, and Y.-H. Hwang, J. Kor. Phys. Soc. 46, 606 (2005). Biotechnol. Prog. 3, 609 (1997). [22] O. R. Fennema, Food Chemistry, New York, Marcel [4] C. Schebor, L. Burin, M. P. Buera, J. M. Aguilera Dekker, 1985. and J. Chirife, Biotechnol. Prog. 13, 857 (1997). [23] F. Sussich, S. Bortoluzzi and A. Cesaro, Thermo- [5] A. Patist and H. Zoerb, Colloids Surf. B: Biointer- chemica Acta 391, 137 (2002). faces 40, 107 (2005). [24] F. Sussich, F. Princivalle and A. Cesaro, Carbohydr. [6] J. L. Green and C. A. Angell, J. Phys. Chem. 93, Res. 322, 113 (1999). 2880 (1987). [25] H. Nagase, T. Endo, H. Ueda and M. Nakagaki, Car- [7] C. A. Angell, Hydrogen-Bonded Liquids(NATO ASI bohydr. Res. 337, 167 (2002). Series C), Kluwer, Dordrecht, 329, 59 (1991). [26] B. S. Chang, R. M. Beauvais, A. Dong and J. F. Car- [8] L. M. Crowe, D. S. Reid and J. H. Crowe, Biophys. J. 71, 2087 (1996). penter, Arch. Biochem. Biophys. 331, 249 (1996). [9] M. K. Dowd, P. J. Reilly and A. D. French, J. Comp. [27] R. Karmas, M. P. Buera, M. Karel and J. Agric. Chem. 12, 102 (1992). Food Chem. 40, 873 (1992).