This work is concerned with the decomposition of dextran by dextranase action (Sugazym DX L, SternEnzym). Detailed analysis of dextran and potential decomposition products is key to this work. Different analytical methods (Haze, Roberts’ Copper Method, Chromatography) were used and new insights into the enzymatic degradation of dextrans could be derived from their combination. The Haze method rather non-specifically detects larger dextrans. This is practical because the larger dextrans are known to predominantly cause processing problems. However, what degradation products appear and how they influence the manufacturing processes is not understood in great detail. The work performed hence actually tries to elucidate this subject matter. Combination of industrially relevant enzyme and dextran levels yielded decomposition products in the range of 40 kDa to trisaccharides. At higher enzyme levels decomposition to smaller saccharides appeared. The crystallization experiments performed indicated that the presence of dextran has a significant influence (crystallization, particle size distribution, crystal morphology). It is found that dextran led to a wider particle size distribution. Light microscopy images illustrate morphological changes induced by the presence of dextran (distinctly elongated, slightly elongated and agglomerated crystals). The precise effect of different dextran fractions on the crystallization is still not fully mapped and certainly needs further investigation. Nevertheless, a negative effect on the crystallization and the crystal characteristics could be detected, while an enzyme treatment reduces or rather eliminates these negative effects taken appropriate enzyme dosages and exposure times are used.
Dextran is an unwanted polysaccharide produced by contaminating microorganisms during sugar processing. This can occur in sugar beet and sugar cane but the climatic conditions in the areas, where sugar cane is cultivated, promote the growth of mesophilic bacteria. Moreover, freezing followed by thawing processes lead to increased dextran levels in sugar beet as well.
The major contributor to this deterioration is the lactic acid bacterium Leuconostoc mesenteroides. The dextran structure can vary according to the strain of the microorganism. Dextran is composed of glucose units, which are mainly linked by -(1à6) glycosidic linkages. Besides, -(1à2)-, -(1à3)-, -(1à4) glycosidic linkages can occur as branching points in the chain (Promraksa, 2008). Dextran formed by L. mesenteroides strains contains approximately 95% of -(1à6)-linkages and 5% branching points (Khalikova et al., 2005).
L. mesenteroides strains produce dextran by enzymatic activity. Dextransucrase enzymes hydrolyse the sucrose molecule to fructose and glucose monomers. The latter is then polymerized to glucose polysaccharides. These polymers are extracted in the mills along with the juice. The presence of dextran in juices/syrups causes several adverse effects during the sugar process as well as on the final product quality. Depending on the molecular weight and the concentration, dextran leads to an increased viscosity of the sucrose solution which affects the filterability, the settling and evaporation as well as the crystallisation rate. Furthermore dextran is supposed to cause a modification of the crystal morphology, which is mainly reported as an elongation along the c-axis (Promraksa, 2008). However, the exact influence of the different dextran fractions on the crystallization is still not completely understood.
The hydrolysis of dextran by enzymes is a promising and often used method to minimize the above mentioned processing problems. Dextran-degrading enzymes are produced by various microorganisms (Khalikova et al., 2005). The hydrolysis of dextran results in a gradual decrease in the average molecular weight. Hence the initial high molecular weight dextran is hydrolysed step by step to lower molecular weight fragments, which in turn are hydrolysed to oligosaccharides, isomaltotriose and isomaltose (Eggleston et al., 2009).
In literature generally a critical value of 500 mg dextran per kg sucrose is reported. Furthermore high molecular weight dextran is reported to be the major contributor to the viscosity increase while low molecular weight dextran is supposed to be the major contributor to crystal modifications (Abdel-Rahman, 2007). The relatively low levels and the wide range of the molecular weight make it difficult to determine the dextran fractions present in a juice exactly (Day & Sarkar, 1986).
In order to better understand the dextran related processing problems the detailed determination and characterization of dextran is a necessary prerequisite. But this still remains a great challenge. The present work concerns the enzymatic decomposition of high molecular weight dextran by dextranase produced by a Chaetomium gracile strain (Sugazym DX L).
The first part of this study includes the determination of the dextran levels and decomposition products by different methods making steps toward a better understanding of the kinetics and mechanism of the hydrolysis process.
In order to assess the benefit of the dextranase treatment it is further necessary to elucidate which specific effects the different dextran fractions have on the sugar production process. Therefore, the second part of this work comprises the investigation of the impact of dextran according to the concentration and molecular weight as well as the effect of the enzymatic treatment of high molecular weight dextran on the crystallization process including the crystal morphology as well as the crystal size distribution.
Material and methods
In the present work the dextran-degrading enzyme Sugazym DX L (SternEnzym) was used to degrade high molecular weight dextran with molecular weights ranging from 1,500,000 to 2,800,000 Da (Sigma-Aldrich, T2000). Furthermore, dextrans of lower molecular weight of 500 kDa (Carl Roth, T500) and 40 kDa (Carl Roth, T40) were used.
15% (w/w) sugar solutions with defined dextran concentrations of 2000 mg and 5000 mg high molecular weight dextran per kg sucrose were prepared and subsequently incubated. The pH-values were around 5.5. The liquid enzymes were dosed precisely and the samples were incubated in a water bath at 65°C for 10 minutes. The incubation time was terminated by thermal inactivation of the enzyme by exposure to 80°C for 20 minutes.
Determination of dextran
After starch removal (amylase treatment) and protein removal (precipitation with trichloracetic acid) followed by filtration, dextran is precipitated with ethanol and the turbidity is measured in a spectrophotometer (720nm) (ICUMSA, 2011).
Roberts’ Copper method
The dextran concentration was determined according to (Roberts et al., 1983) and (Khaleifah, 2001). All of the polysaccharides are precipitated with ethanol and the dextran is then selectively precipitated with alkaline copper sulfate, which is then determined colourimetically forming a phenol-sulfuric acid colour complex in a spectrophotometer (485nm).
Dextran level was then calculated with the following equation:
A – wt. of sample solids diluted to 100 ml
B – ml of aliquots taken for alcohol precipitation
C – ml of solution of alcohol precipitate
D – ml of aliquot taken for copper precipitation
E – ml of final solution of copper-dextran complex
F – mg/ml dextran received from standard curve
Realistic dextran levels are below the detection limit of the used gel permeation chromatography (GPC) and high performance liquid chromatography (HPLC). To allow a detailed analysis, the enzyme treatment was done with concentrated juices, scaling both, the enzyme and the substrate level by a factor 40.
Performance of the crystallization
Laboratory crystallisations were performed as described elsewhere (Schlumbach et al., 2015). Evaporative crystallization was realized in a 5 liter crystallizer. The process is controlled such that at practically constant supersaturation approximately 50% crystalline material is generated. Also separation/affination, and drying are executed such that industrial relevance is ensured. A fluidized bed was used to dry the sugar.
The different feedstocks were prepared as 65% (w/w) sucrose solution with dextran additions according to Table 1. The crystallization of enzyme treated sugar solutions required a prior evaporation of the 15% (w/w) sugar solution to approximately 65-67% (w/w) sugar solution. Enzyme concentrations of 4 mg and 60 mg per kg juice were chosen based on the prior analytical results.
Table 1 Dextran and enzyme concentrations of the crystallization experiments
|Dextran average molecular weight [kDa]||—||40||40||40||500||2000||2000||2000||2000||2000|
|Dextran conc. [mg/kg sucrose]||0||5000||2000||500||2000||5000||2000||500||5000||5000|
|Enzyme conc. [mg/kg juice]||4||60|
Analysis of the sugar crystals
The analysis of the particle size distribution by sieving was done according to the ICUMSA Method GS2/9-37 (ICUMSA, 2011). Furthermore, light microscopy images were recorded by Zeiss-Axio Scope A1 microscope and subsequently analysed by ImageJ.
Results and discussion
Determination and characterization of dextran in sucrose solutions after enzyme treatment
The Haze and the Roberts’ Copper methods
The existing methods available have different detection limits. The Haze Method is the industrially commonly applied dextran determination method because of the rapid and easy experimental procedure. The main disadvantage of this method is its limitation to high molecular weight dextran (Rauh et al., 1999). This method is not capable of determining low molecular weight dextran smaller than 40 kDa (Abdel-Rahman, 2007).
In contrast the Roberts’ Copper Method is more complex and time-consuming. It determines the integral concentration of dextrans bigger than trisaccharides. It does however not allow for distinction between the different dextran fractions (Roberts et al., 1983).
Hence neither the Haze method nor the Roberts’ Copper methods are capable of identifying decomposition products. Nevertheless, the combination of both methods is useful to determine the amount of high and low molecular weight dextran by simple calculation based on the knowledge of the different determination limits.
Figure 1 shows the dextran levels determined by the Haze and the Roberts’ Copper method of samples with initial dextran levels of 2000 mg dextran/kg sucrose before and after enzyme treatments. Variation of the enzyme concentration was investigated. A critical value of 500 mg dextran/kg sucrose was assumed based on results of others (Abdel-Rahman, 2007).
The second set of bars from left shows the remaining dextran levels determined after enzyme treatment with the initial parameters 2 and 4 mg enzyme/kg juice. The Haze method indicates an almost complete decomposition with dextran levels below 300 mg/kg sucrose.
In contrast the Roberts’ Copper values indicate significantly higher dextran values. The application of enzyme concentrations of 2 and 4 mg/kg juice lead to dextran levels of 1558 mg/kg sucrose and 1003 mg/kg sucrose. Obviously, there is a great difference between the dextran levels determined. The Haze method indicates an almost complete degradation of dextran at relatively low dextranase levels, which illustrates the removal of high molecular weight dextran possibly eliminating the viscosity problem. On the other hand, the Roberts’ Copper Method indicates high dextran levels due to its integration of a wide range of molecular sizes which probably overestimates the dextran level.
The first set of bars from left shows the dextran level without an enzyme treatment determined by both methods, which also indicates that there are little analytical imprecisions for these two methods. The Roberts’ method shows slightly higher while the Haze method shows slightly lower dextran values than expected. This indicates that the Roberts’ Method seems to slightly overemphasize the dextran issue not only because of the integration of all molecular sizes. The opposite applies for the Haze Method. This analytical inaccuracy also slightly contributes to the great difference between the dextran levels determined.
Nevertheless, the combination of these methods is useful to get an idea of the molecular weight distribution present in the juice after enzyme treatments. The great difference at these enzyme levels suggests the presence of a relatively high proportion of molecules in the range of low molecular weight to trisaccharides. As described above dextrans of lower molecular weight have a significant influence as well and should hence be eliminated. This can for example be achieved by higher enzyme dosage with 8 to 16 mg enzyme/kg juice (third set of bars from left in Figure 1). It is well known that the substrate concentration has a strong influence on the enzyme kinetic (Michaelis-Menten). Figure 2 shows analogue to Figure 1 the dextran levels determined by the Haze and the Roberts’ Copper method after different enzyme treatments of juices with an initial dextran value of 5000 mg/kg sucrose. A very similar principle as shown in Fig. 1 emerges. Again, a great difference between the values is noticeable, indicating the presence of a high proportion of these decomposition products in the range of low molecular weight to trisaccharide. And again, much higher enzyme concentrations (60 mg enzyme/kg juice) are necessary to degrade the low molecular weight dextran to the critical value of 500 mg dextran/kg sucrose (third set of bars from left in Figure 2).
The further increase from 25 mg/kg to 40 mg/kg juice (Figure 2) does not yield a reduction in the dextran levels. This is possibly due to the fact that within this processing window the size of the dextrans is reduced but no decomposition to small fragments such as trisaccharides yet occurs. In other words the Roberts’ Copper method is not capable of monitoring the size reduction of dextran unless molecules with three or less glucose units evolve (Roberts et al., 1983).
This also illustrates that most likely the single usage of both methods are not suited to indicate dextran problems in detail. The Haze method is certainly too inaccurate but Roberts’ Copper method taking all molecules larger than trisaccharides into account might possibly overemphasise dextran problems. However, the combination of the results allows a first assessment of the decomposition process by distinguishing, high, low and decomposed dextrans. The medium to low molecular weight dextrans are hence the explanation for the massive differences obtained with the two analytical methods.
Figure 1 Dextran concentration [mg/kg sucrose] after enzyme treatment in 15% (w/w) sucrose solutions with an initial dextran value of 2000 mg/kg sucrose; variation of the enzyme level (65°C, 10 min; pH:5.5)
Figure 2 Dextran concentration [mg/kg sucrose] after enzyme treatment in 15% (w/w) sucrose solutions with an initial dextran value of 5000 mg/kg sucrose; variation of the enzyme level (65°C, 10 min; pH:5.5)
For a more detailed determination, gel-permeation-chromatography (GPC) and high-performance liquid chromatography (HPLC) were used. In addition to the wide range of the molecular weight, the relatively low dextran levels make it difficult to determine dextran exactly (Day & Sarkar, 1986).
This became apparent using the previous dextran levels. No peaks for dextran or other components except of sucrose could be detected by GPC and HPLC. As a result, it was necessary to find a way to concentrate the samples. In this case, the enzyme as well as the substrate concentrations were augmented (by factor 40).
The red curve in Figure 3 (A) presents a 15 % (w/w) sugar solution containing 80,000 mg dextran/kg sucrose of high molecular weight dextran (T2000-2000 kDa), which is the initial dextran for the enzymatic hydrolysis. For largest elution volumes (latest elution) the 15 % (w/w) sucrose causes a huge peak. In general, the chromatograms indicate that the dextran samples show a rather wide size distribution with T2000 clearly distinguishable (Figure 3 (B)). The green curve illustrates that the three fractions (T2000, T500, T40) clearly overlap. As expected, the larger molecular weights elude first.
Obviously, the T2000-peak of the green curve is higher in comparison to the red one although both of them have the same concentration of high molecular weight dextran (T2000). Consequently, it is probable that T500-dextran contributes to this peak. The second dextran peak of the green curve seems to represent T500- and T40-dextran. Despite of this overlapping effect, the end of low molecular weight dextran (T40) is clearly identifiable, which is marked by the return to the basic line.
In Figure 3 (C) the chromatograms for enzyme treated samples are shown (excluding the sucrose peak). The exaggerated dextran (T2000) concentrations of 80,000 mg/kg and 200,000 mg/kg sucrose prior to the enzymatic treatment were used. In analogy to the dextran concentration also the enzyme concentration was raised by a factor 40 for these experiments.
The dextranase action leads to a peak development between or rather slightly overlapped with the T40-peak and the sucrose peak, which indicates the presence of low molecular weight to oligosaccharide products. This is in agreement with the conclusion assumed by the Haze and the Roberts’ Copper method for these enzyme to substrate ratios.
Figure 3 (A) GPC-chromatograms of all samples (B) dextran peak characterization (C) decomposed dextran
Red curve: 80,000 mg dextran (T2000)/kg sucrose Green curve: Dextrans T40, T500, T2000 each 80,000 mg/kg s. Black curve: T2000: 80,000 mg/kg sucrose treated with 80 mg enzyme/kg j. Yellow curve: T2000: 80,000 mg/kg sucrose treated with 160 mg enzyme/kg j. Purple curve: T2000: 200,000 mg/kg sucrose treated with 80 mg enzyme/kg j. Blue curve: T2000: 200,000 mg/kg sucrose treated with 160 mg enzyme/kg j.
A higher substrate as well as a higher enzyme concentration leads to a higher enzyme-substrate-contact which increases the enzyme reaction rate (Michaelis-Menten). Apparently, the higher substrate concentration leads to a higher amount of decomposition products without changing the size of the decomposition products, as comparison of the purple and black line reveal. Although the enzyme concentration is the same for each of the two lines, higher concentrations of decomposition products were observed for the higher substrate concentration. The peaks of the measured curves are in alignment (dotted line in Figure 3 (C)), indicating the same size range. Consequently, it seems like the enzyme is not saturated at this enzyme to substrate ratio, which was not indicated by the results obtained by the Haze and the Roberts’ Copper method. This can be explained by the change of the enzyme kinetic due to the multiplication of the substrate and enzyme concentration, which results in a higher enzyme-substrate-contact.
The curve for the higher enzyme level elutes slightly later and shows a slightly smaller peak. Consequently, the increase of the enzyme concentration leads to smaller decomposition products at a lower amount indicating more fully decomposed dextran (demonstrated by the horizontal arrow in Figure 3 (C)).
For these four artificially highly dextran loaded juices treated with dextranase it is found that the dextran fragments resulting from the decomposition are rather small. Their molecular weight clearly below 40 kDa indicates the presence of a substantial amount of oligosaccharides.
The HPLC column used was able to measure mono-, di- and trisaccharides. Again, at the low concentrations of the initial samples HPLC could not detect any of the components. As a result, the samples with the increased enzyme and substrate concentration were used. These chromatograms indicate a peak development in the range of mono-, di-, trisaccharides. The peak with the highest elution volume (RT 26 min) represents the sucrose peak. The decomposition product peak elutes slightly later (RT 36 min). Known retention times help to identify the developed peak, indicating that the developed peak mainly belongs to isomaltose.
In addition to that, it is also worth to consider disaccharides evolved through the existing branching points in the dextran structure. As mentioned before, dextran produced by L. Mesenteroides contains about 95% α-1,6 glycosidic bonds and 5% α-1,3-glycosidic bonds. Hence the presence of a disaccharide made of two glucose units linked by α-1,3-glycosidic bonds (nigerose) could be possible as well, which potentially contributes to this peak. The enzyme acts specifically on α-1,6-bonds and therefore does not cleave α-1,3-bonds. However, it is questionable whether the concentration present is sufficient to detect by the used HPLC.
And again, the influence of the enzyme as well as of the substrate concentration is obvious and illustrated in Figure 4 (A). Higher substrate as well as enzyme concentrations lead to an increase of the peak height, which is proportional to the amount of the decomposition product.
Figure 4 HPLC-Chromatograms of the enzyme treated samples Red curve: 80,000 mg dextran (T2000) /kg sucrose
Black curve: T2000: 80,000 mg/kg sucrose treated with 80 mg enzyme/kg j. Yellow curve: T2000: 80,000 mg/kg sucrose treated with 160 mg enzyme/kg j. Purple curve: T2000: 200,000 mg/kg sucrose treated with 80 mg enzyme/kg j. Blue curve: T2000: 200,000 mg/kg sucrose treated with 160 mg enzyme/kg j.
Decomposition products in summary
The above explained results indicate the presence of decomposition products mainly in the range of oligosaccharides at relatively low enzyme levels. A smaller part is presented by smaller molecules (mainly isomaltose), which increases with a higher enzyme level. This could be confirmed by chromatographic measurements.
In order to assess the benefit of the dextranase treatment it is further necessary to elucidate which specific effects the different dextran fractions as well as potential decomposition products have on the sugar production process. A possible negative impact of decomposition products should be reduced or eliminated. Therefore, the next chapter concerns the influence of dextran and the gain of an enzyme treatment with regard to the crystallization process.
Impact on crystal morphology
Impact of dextrans according to the molecular weight and concentration
Light microscopy images were analysed using ImageJ. Two-dimensional crystals were analysed excluding the third axis. The crystals probably lay on the 100-surface of a sucrose crystal which means the software analyses the c- and b-axis. Two parameters, the circularity and the aspect ratio, were measured and the relative frequencies were calculated and illustrated as a distribution curve.
The Aspect Ratio is the proportion of the major to the minor axis of an ellipse placed in the sugar crystal. For a distinct elongation of a crystal, the aspect ratio is supposed to increase. On the contrary, a slight crystal elongation along the c-axis leads to a decrease of the aspect ratio due to the approach of the two axes. Figure 5 (A) shows the relative frequencies of the samples with the highest dextran concentration (5000 mg/kg sucrose), the reference and the enzyme treated samples. A shift of the distribution curve to smaller aspect ratios is found for the samples which are grown in the presence of dextran. Consequently, there is a higher amount of crystals with a lower aspect ratio in comparison to the reference, which could indicate the presence of more crystals with a slight elongation along the c-axis.
Figure 5 Relative frequencies (A) Aspect ratio (B) Circularity
The second important parameter for the image analysis is the circularity, whereat a perfect circle has a value of 1.0. A decrease of the circularity indicates a distinct elongation of the crystal, resulting in needle-shaped crystals. Slightly elongated crystals and therefore lower values for the aspect ratio show an increased circularity due to the approach to a quadratic shape. However, the relative frequencies of the circularity also indicate a higher amount of crystals with a lower circularity, as shown in Figure 5 (B).
Figure 6 shows an exemplified selection of the different possible sucrose crystal shapes and correspondingly values for the circularity and the aspect ratio. Crystal (A) represents a reference crystal.
Besides there are some crystals in Figure 6 which are grown in the presence of 5000 mg/kg sucrose of T2000- and T40-dextran. As mentioned before, this leads to a distinct elongation, hence an increase in the aspect ratio and a decrease in the circularity, as represented by Crystals (C) and (D). Crystal (B) illustrates the approach of the c-axis to the b-axis resulting in a lower aspect ratio and a higher circularity. Furthermore, the presence of dextran seems to increase the agglomeration processes. Crystal agglomerates show a low value for the circularity while the aspect ratio is form-related slightly higher or lower or rather unchanged.
Table 2 summarizes the particular effects of the identified crystal shapes on the values of the circularity and the aspect ratio. With the help of these analysed isolated crystals in Figure 6, it is known that these crystals occur. Two of the given crystal types show a decrease of the circularity which could explain the lower total value of the circularity. Similar considerations apply for the aspect ratio.
Looking at the totality of all values, it is likely that the main part of the crystals was unaffected by the dextran available. Potential crystal changes are probably composed of rarely appearing distinctly elongated crystals (C) and (D), slightly elongated crystals (B) and agglomerated crystals (E) and (F).
Table 2 Effect of the circularity and the aspect ratio according on the particle shape
|Particle shape||Circularity||Aspect Ratio|
|Slightly elongated crystal||increase||decrease|
Effect of the enzyme treatment
Figure 5 (B) show that the enzyme treatment with 4 and 60 mg enzyme/kg juice leads to a similar progression of the distribution curve for the circularity of the reference.
The relative frequencies for the aspect ratio for the sample with the lower enzyme concentration still indicate the presence of particles with a lower aspect ratio in comparison to the reference. With the help of Table 2, one can assume a reduction or even elimination of distinctly elongated crystals as well as of agglomerated particles.
The increase of the enzyme concentration to 60 mg enzyme/kg juice leads, just as the values for the circularity, to an approach of the distribution curve of the aspect ratio to the reference. Consequently, the lower enzyme concentration and therefore the decomposition to molecules mainly in the range of 40 kDa to trisaccharides, leads to an improvement of the crystal morphology, but it does not completely eliminate it.
A further enzymatic decomposition due to a higher enzyme level shows distinctly better crystal shapes resembling the reference (not only in circularity, but also in aspect ratio).
Hence the slightly larger decomposition products developed by the application of a lower enzyme concentration seem to still have a little effect on the morphology, whereas distinctly elongated and agglomerated crystals are probably reduced or rather eliminated. The smaller decomposition products, which prevail at higher enzyme concentration, does not show an effect on the crystal shape in terms of all measured parameters.
Particle size distribution
The sieve analysis was used to investigate the particle size distribution. The coefficient of variation is shown in Figure 7 (A) . It is apparent that the presence of dextran significantly impacted crystal size distribution. High as well as low molecular weight dextran cause an increase of the coefficient of variation and a decrease of the mean aperture. This indicates a wider particle size distribution, embracing the production of small-sized crystals. The calculation of the percentages of the particles retained on each sieve supports this assumption (Figure 7 (B)). Apparently, there is a shift of the particle size distribution curve towards smaller aperture sizes.
The presence of a higher amount of smaller crystals could be explained by a function of dextran as seed crystals. The decrease of the interfacial energy and therefore of the critical radius due to foreign particles results in a higher nucleation rate. So it could be possible that dextran has an influence on the interfacial energy.
The effect was highest for high molecular weight dextran. All of the used dextrans originate from L. mesenteroides, which have a similar distribution of glycosidic bonds. Hence the single factor contributing is the molecular size, which may increase the probability to adsorb at the crystal surface.
The effect on an enzyme treatment on the particle size distribution is also shown in Figure 7. An enzyme concentration of 4 mg/kg juice leads to an improved coefficient of variation close to the reference and shows an approach of the particle size distribution curve. The further increase of the enzyme concentration results in a distinct improvement of the coefficient of variation, even better than the reference, and to a more improved particle size distribution curve.
Figure 7 Coefficient of variation of the particle size distribution by sieving; ●-Reference ; ▲-T40; ■-T500; ♦-T2000; ×-4 mg/kg juice; + 60 mg enzyme/kg juice
In the current research the enzymatic decomposition of dextran was analysed by the determination and characterization of remaining dextran and potential decomposition products. Depending on the enzyme level, decomposition products in the range of 40 kDa to trisaccharides and molecules smaller than this were found. The presence of high as well as low molecular weight dextran has an influence on both the crystal morphology and the crystal size distribution. However, to detail the effect of dextran on the crystallization still needs some further investigation. Nevertheless, a negative effect on the crystallization and the crystal characteristics was detected, which can be mitigated or even eliminated by enzyme treatment.
Abdel-Rahman, E. (2007). Investigation on the influence of dextran during beet sugar production with special focus on crystal growth and morphology. PhD. Berlin.
Day, D. F., & Sarkar, D. (1986). Methods of Analysis for Dextran in Sugar Molasses and Juice. Proc. Int. Soc. Sugar Cane Technol, 748–754.
Eggleston, G., Monge, A., Montes, B., & Steward, D. (2009). Application of dextranase in sugarcane factory: Overcoming practical problems. Sugar Tech, 11, 135–141.
ICUMSA (Ed.) (2011). ICUMSA Method Book. Berlin: Dr. Albert Bartens KG.
Khaleifah, M. (2001). Effect of Dextran on Sugarcane Quality and Raw Sugar Manufacture. PhD. Assiut.
Khalikova, E., Susi, P., & Korpela, T. (2005). Microbial dextran-hydrolyzing enzymes: fundamentals and applications. Microbiology and molecular biology reviews MMBR, 69, 306–325.
Promraksa, A. (2008). Reduction of Dextran in Raw Sugar Production. PhD. Suranaree.
Rauh, J. S., Cuddihy J.A., & Opelka, M. J. (1999). Analyzing Dextran in the Sugar Industry: A review of dextran in the factory and a new analytical Technique. American Society of Sugar Beet Technologists, 30, 29–40.
Roberts, E. J., Clarke, M. A., & Godshall, M. A. (1983). The analysis of dextran in sugar production. Proc. Int. Soc. Sugar Cane Technol, 18, 1374–1382.
Schlumbach, K., Pautov, A., Göckeritz, L., Bagherzadeh, A., & Flöter, E. (2015). Controlled sucrose crystallization at pilot-plant scale. Sugar Industry, 140, 500–507.