Biotechnology Letters

, Volume 36, Issue 3, pp 531–536

Production of fructose from highly concentrated date extracts using Saccharomyces cerevisiae


  • Meilana Dharma Putra
    • Department of Chemical Engineering, College of EngineeringKing Saud University
    • Department of Chemical Engineering, College of EngineeringKing Saud University
  • S. M. Al-Zahrani
    • Department of Chemical Engineering, College of EngineeringKing Saud University
  • M. H. Gaily
    • Department of Chemical Engineering, College of EngineeringKing Saud University
  • A. K. Sulieman
    • Department of Chemical Engineering, College of EngineeringKing Saud University
  • M. A. Zeinelabdeen
    • Department of Chemical Engineering, College of EngineeringKing Saud University
  • H. K. Atiyeh
    • Biosystems and Agricultural EngineeringOklahoma State University
Original Research Paper

DOI: 10.1007/s10529-013-1388-y

Cite this article as:
Putra, M.D., Abasaeed, A.E., Al-Zahrani, S.M. et al. Biotechnol Lett (2014) 36: 531. doi:10.1007/s10529-013-1388-y


Large amounts of low-quality dates produced worldwide are wasted. Here, highly concentrated fructose syrups were produced via selective fermentation of date extracts with Saccharomyces cerevisiae. Syrups with 95.4–99.9 % (w/w) fructose yields were obtained from date extracts having an initial sugar range of 49–374 g/l without media supplementation; the corresponding ethanol yields were between 69 and 52 % (w/w). At 470 g initial sugars/l, fructose and ethanol yields were 84 and 47 % (w/w), respectively, and the product contained 62 % (w/w) fructose, which is higher than the widely available commercial 42 and 55 % (w/w) high fructose corn syrups. The commercial potential for conversion of waste dates to high-value products is thus demonstrated.


DatesEthanolFructoseHigh fructose syrupSaccharomyces cerevisiaeSelective fermentation


Palm dates (Phoenix dactylifera) are grown in many parts of the world and in 2010 the worldwide production was about 7.9 million tons (Jain 2012). Unfortunately, most of the dates produced are wasted (Moshaf et al. 2011). Dates contain mainly equal amounts of fructose and glucose with small amounts of sucrose, protein, minerals and vitamins (Baliga et al. 2011). Fructose is the sweetest natural sugar, 30 % sweeter than sucrose and 80 % sweeter than glucose. Because they are isomers, glucose and fructose are very difficult to separate when present in a mixture. Fructan-rich natural materials, such as Jerusalem artichoke, chicory and dahlia, tubers can be hydrolyzed to produce fructose (Abasaeed and Lee 1995), which is widely used in the food, beverage, confectionery and pharmaceutical industries (Johnson et al. 2009). It is, therefore, critically important to find new ways to use the natural, sustainable and unused resource of dates.

Because of equilibrium limitations, industrial enzymatic isomerization processes produce high fructose syrups (HFS) that contain only 42 % (w/w) fructose (Gaily et al. 2010; Zhang et al. 2004). 90 % (w/w) HFS can be produced via multistage chromatographic processes or membrane technology (Paugam et al. 1996) but both processes suffer from high cost and separation difficulties. A promising process for the production of fructose is selective fermentation of sugars to ethanol (Koren and Duvnjak 1992). Preliminary economic analysis has shown the viability of this process (Carvalho et al. 2008). Using high initial sugar concentration for the production of ethanol minimizes subsequent distillation costs and avoids osmo-sensitive contaminants in the fermentation broth (Jones et al. 1981).

Many microbial strains have been used to ferment sucrose media or glucose/fructose mixtures (Atiyeh and Duvnjak 2001). However, a fermentation process using most of these strains showed significant losses of fructose (up to 50 % w/w in some cases) and formation of undesired by-products, e.g., sorbitol, >4 % (w/w) (Carvalho et al. 2008). Sacharomyces cerevisiae showed high fructose and ethanol yields (Atiyeh and Duvnjak 2001).

Co-production of fructose and ethanol from date extracts at high initial sugar concentrations (>140 g/l) is challenging. Therefore, the objective of the present study was to examine the ability of S. cerevisiae to produce high fructose syrups and ethanol from highly concentrated date extracts without nutrient addition during fermentation.

Materials and methods

Raw materials

Sugars were extracted from dates using deionized water at 40 °C for 2 h. The ratio of pitted dates to water was 2:5 (w/w). The extract was centrifuged at 9500×g for 6 min to remove suspended solids and fibers. The final date syrup, without supplement addition, was then autoclaved at 121 °C for 15 min. The initial sugar concentration (ISC) in the prepared date syrups was between 49 and 470 g/l.

Microorganism and media

The inoculum was prepared by aseptically transferring a loopful of S. cerevisiae ATCC 36858 from an agar slant to a 500 ml Erlenmeyer flask containing 100 ml medium. The Yeast Malt Broth used consisted of 3 g bacto-yeast extract, 3 g bacto-malt extract, 5 g bacto-peptone and 10 g bacto-dextrose and de-ionized water (up to 1 l). The medium was then autoclaved. The transferred yeast was further propagated for 48 h at 30 °C with shaking.

Fermentation process

Fermentation experiments were carried out in 500 ml Erlenmeyer flasks (100 ml working volume) placed in a rotary shaker at 30 °C. Liquid samples were withdrawn periodically to measure cell mass, sucrose, fructose, glucose, ethanol, glycerol and sorbitol concentrations. One set of experiments was conducted in a 1 l fermentor with an ISC of 139 g/l for comparison with the results of the 500 ml flasks.

Analytical procedures

Cell mass concentration (biomass), was determined using the dry weight method. Samples were dried overnight at 105 °C. Cell count and viability were determined using a cell counter system.

Sugars, ethanol, glycerol and sorbitol concentrations were determined using HPLC equipped with an RI detector and Aminex column. The column was at 40 °C and 1 mM H2SO4 was used as the mobile phase at 0.8 ml/min.

Results and discussion

Selective fermentation of date syrups

Figure 1 shows the kinetic profiles of glucose-selective fermentation of date syrups in the 1 l fermentor. The yeast grew on glucose during the early stages of the fermentation, followed by production of ethanol that continued to increase towards the end of the process. Almost all of the glucose was consumed after 28 h. The fructose yield (defined as the ratio of the amount of fructose at any time to its initial amount) at 28 h was 94.7 % (w/w) and the ethanol yield was 77 % (w/w) of its theoretical value (calculated based on the consumption of total sugars). During the first 28 h of fermentation, the biomass increased, giving a biomass yield of 0.034 g/g of consumed sugars. The pH dropped from 4.9 to 4.2 through production of acids (Stewart and Russell 1987). About 1.7 g glycerol/l was produced at 28 h and this slightly increased to 1.9 g/l after 72 h. Sucrose dropped by 10 % (w/w), confirming the ability of this strain to hydrolyze sucrose (Atiyeh and Duvnjak 2001). Fructose yield dropped upon continuation of fermentation beyond 28 h.
Fig. 1

Profiles of various components and pH during the fermentation of date syrup having an initial sugar of 139 g/l in the 1 l fermentor

Effect of initial sugar concentrations in date syrups

Table 1 summarizes the effect of ISC in the date syrups on the production of biomass, fructose and ethanol. Fructose yield decreased by 18 % (w/w) at higher ISC. In contrast with the high selectivity of S. cerevisiae, almost half of the fructose was consumed when using Zymomonas mobilis (Bringer-Meyer et al. 1985; Doelle and Doelle 1991). Less than 20 % (w/w) of fructose was consumed when using S. cerevisiae to ferment a substrate containing Jerusalem artichoke extract supplemented with synthetic glucose (Koren and Duvnjak 1992).
Table 1

Performance of S. cerevisiae in producing fructose, ethanol and biomass from date syrups at various initial sugar concentrations

Initial sugar (g/l)

Initial biomass (g/l)

Biomass yielda (g/g)

Fructose yieldb (%) (w/w)

Ethanol productivitya (g l−1 h−1)

Ethanol yielda (%) (w/w)

















































aCalculated at maximum ethanol concentration

bCalculated at minimum glucose concentration

cProcessed in the 1 l fermentor

Significant drops in ethanol yield and productivity were observed with increasing ISC. At the lowest ISC, the ethanol yield was 52 % (w/w) of its theoretical value because of the relatively high glucose consumption during the short lag phase, as manifested by a much higher biomass yield compared with those obtained at 139 g/l. The biomass yield dropped with increasing ISC due to substrate inhibition (Jones et al. 1981). Ethanol yields obtained in the present study were lower than those obtained by Atiyeh and Duvnjak (2001) but were slightly higher than those obtained by the same authors (Atiyeh and Duvnjak 2002) at high ISC.

For comparative purposes, identical experiments at 139 g initial sugar/l were conducted in a 1 l fermentor. Slightly better results were obtained than for experiments performed in the flasks. This slight improvement was attributed to better mixing and also to the continuous and uninterrupted operation of the system when taking samples (Nuanpeng et al. 2011).

Figure 2a shows glucose consumption profiles during fermentation at various ISC. Sugar consumption was slow at 470 g/l, with very little production of ethanol. S. cerevisiae was shown to endure higher carbon source of 726.4 g sucrose/l (Atiyeh and Duvnjak 2001) compared with Z. mobilis, which endured 462 g/l (Doelle and Greenfield 1985). At higher ISC (>360 g/l), S. cerevisiae hydrolyzed sucrose rather than fermenting glucose. Production of fructose from glucose/fructose mixtures having ISC higher than 415 g/l was reported by Koren and Duvnjak (1992). Also, Fig. 2a shows that only half of the glucose was consumed after 120 h because of ethanol inhibitions (Lee et al. 1979). Figure 2b reveals a low consumption of fructose, even when glucose was almost completely consumed.
Fig. 2

Concentration profiles of (a) glucose; (b) fructose at various initial sugar: [49, filled diamond; 139, filled square; 250, filled triangle; 290, filled circle; 374, open square; 444, open circle; 470, open triangle (g/l)]. The embedded small box compares the results at 139 g/l in the 1 l fermentor and a 500 ml flask

Figure 3a shows that a higher ISC led to a lower growth of yeast. In addition, the biomass was above 4.5 g/l at ISC below 374 g/l. The drop in biomass during the first 24 h for ISC of 470 g/l could be attributed to the extreme osmotic state imposed on the yeast (Jones et al. 1981). As expected, Fig. 3b shows that higher ethanol concentrations were obtained at higher ISC. However, the fermentation process required more time, thus resulting in lower ethanol productivity (Table 1).
Fig. 3

Concentration profiles of (a) biomass; (b) ethanol production at various initial sugar: [49, filled diamond; 139, filled square; 250, filled triangle; 290, filled circle; 374, open square; 444, open circle; 470, open triangle (g/l)]. The embedded small box compares the results at 139 g/l in the 1 l fermentor and a 500 ml flask

Fructose and by-products

As shown in Table 2, the fructose fraction (i.e., the amount of fructose per amount of total carbohydrates) at the end of the process decreased with increasing ISC. The highest fructose fractions (>90 % w/w) were obtained at ISC below 374 g/l. It is interesting to compare the results obtained at 290 g/l with those obtained at 374 g/l. While the fructose losses were similar, the fructose fractions shown in Fig. 4 were different. Compared with Atiyeh and Duvnjak (2001), the fructose fractions obtained here were lower because of the sucrose content at the end of fermentation. For example, for ISC of 250 g/l, the sucrose, initially 1.5 % (w/w), increased to 2.9 % (w/w) towards the end of fermentation because of fermented glucose. The very slow uptake of sucrose (whereby only 15.4 % w/w of the initial amount was hydrolyzed) could be attributed to its low initial content, causing the strain to ferment glucose rather than hydrolyze sucrose. This finding agrees with the report that in sucrose-rich media, the rate of hydrolysis of sucrose is faster than glucose conversion; however, the opposite is true at low sucrose content (Atiyeh and Duvnjak 2001, 2002). Another factor that could plausibly affect the composition of the products is the supplement content in the media (Maiorella et al. 1984).
Table 2

Performance of S. cerevisiae in consumption of sucrose and production of fructose and by-products during the selective fermentation process

Initial sugar (g/l)

Fructose lossa (%) (w/w)

Fructose fraction (%) (w/w)

Sucrose hydrolyzedb (%) (w/w)

Glycerola (g/l)

Sorbitolb (g/l)

















































aCalculated at minimum glucose concentration

bCalculated at 120 h

cProcessed in the 1 l fermentor
Fig. 4

Profiles of fructose fraction during the fermentation of date syrups at various initial sugar: [49, filled diamond; 139, filled square; 250, filled triangle; 290, filled circle; 374, open square; 444, open circle; 470, open triangle (g/l)]

The economic viability of any process is affected by formation of by-products, because of downstream separation costs. One of the shortcomings of the production of fructose by selective fermentation is the formation of sorbitol. In processes using Z. mobilis ATCC 39676 (Doelle and Greenfield 1985), ATCC 53431 (Bringer-Meyer et al. 1985) and ATCC 53432 (Suntinanalert et al. 1986), at least 35, 60 and 86 g sorbitol/l was formed, respectively. In the present work, only a small amount (<2 g/l) of sorbitol was formed after complete consumption of glucose.

Glycerol increased with ISC but the amounts of glycerol produced were still lower than those reported: e.g., at an ISC of about 270 g/l, processes using media comprising date syrup (290 g/l in Table 2), sucrose [271 g/l in Atiyeh and Duvnjak (2001)], and sugar beet molasses [242 g/l in Atiyeh and Duvnjak (2002)] resulted in glycerol of 4.8, 7.2 and 16.1 g/l, respectively. More glycerol was found in beet molasses because of its higher mineral content (Maiorella et al. 1984). The presence of glycerol affects ethanol yield (Oura 1977) and, because glycerol increased (Table 2) with ISC, the ethanol yield decreased (Table 1).

In conclusion, unused dates constitute a natural and sustainable raw material for the production of high fructose syrups (>90 % w/w) via selective fermentation of the glucose component in date extracts by S. cerevisiae.


The authors extend their appreciation to the Deanship of Scientific Research and the Research Center at the College of Engineering, King Saud University, for supporting this work.

Copyright information

© Springer Science+Business Media Dordrecht 2013