CA2034475C - Method for crystallizing anhydrous fructose from its aqueous solutions - Google Patents

Abstract

Crystallizing anhydrous fructose by preparing an aqueous solution containing at least about 90% dry substance, the fructose content of the dry substance being at least about 90% by weight; seeding the aqueous solution at a temperature of 50-60°C; and cooling the resulting mass at a controlled rate and with continuous mixing such that the supersaturation of the liquid solution with respect to saturated fructose is less than a ratio of about 1.25 and the temperature difference between the warmest and coolest portions of said solution is no more than about 6°C. Cooling is accomplished in a crystallizer with a heat transfer surface of 5 mm2/mm3 or greater. This method enables large scale, high capacity, high yield production of crystalline fructose through the use of a crystallizer with optional heat and mass transfer properties.

Description

IMPRO'lED METHOD FOR CRYSTALLIZING
ANH'!DROUS FRUCTOSE FROM ITS AQUEOUS SOLUTIONS
~.
INTRODUCTION
? FIELD OF THE INVED1TION
3 This invention relates to the production of crystalline fructose. riiore specifically, it provides a method for large scale, high capacity, high yield production of cr~_rstalline fructose through the use of a crystallizes with optimal heat and mass transfer properties.
BACKGROUND OF THE INVENTION
The present invention relates to an improved method for crystallizing anhydrous fructose crystals from water solution. Disclosed herein is an economic method for producing =' large scale, high yields of crystalline fructose. Crystalline '-3 fructose is generally obtained by seeding supersaturated fructose solutions to induce crystall'_ne'growth. Due to the solubility and stability characteristics of fructose and high 16 viscosity of fructose solutions, however, it is often problematic to maintain the optimum conditions to insures the 13 economic production of a pure crystalline product.
~a Fructose is very soluble in water and the solutions 1'0 : are extremely viscous. A large amount of heat due to high '1 crystallization heat of fructose, mixing heat and additional 22 cooling of the mass must be removed. during fructose '3 crystallization. In addition, because fructose has a very narrow metastable zone, the temperature difference between the '' solution and the cooling surface must be quite low thus making '° the crystallisation very difficult.
?7 To overcome this difficulty, several prior art processes involve the use of organic solvents to crystallize fructose from aqueous solutions. In Finnish Patent Application 30 No. 862025, for example, a continuous fruc'~ose crystallization ~' ..
1 ~~ method using organic solvents is described. The viscosity of i the fructose sol.~tion, however, results in a lowering of 3 productivity, thus the yield is only about 40a and the 4 productivity about 0.17 t/m3/d even if the mass is pumped through a vertical crystallizer. The productivity (t/m3/d) is defined as the production rate of crystals (metric tons) per 7 the total volume of the crystallizer (cubic meter).
8 Crystallization from an organic solvent or water solvent mixture is also described in Staley's European patent 015017. The use of organic solvents, however, creates 11 disadvantages with large scale crystallizations. These include 12 fire hazards as well as the fact that solvents are toxic and 13 therefore unsuitable because small residues remaining in the 14 crystalline product will leave it unsuitable for use in foods.
1' Several methods have been developed which avoid the is use of organic solvents in the fructose crystallization 17 process, but these methods are often disadvantageous 18 economically because of the high viscosity and unstable nature 19 , of superstaturated fructose solutions. UK Patent Application i 2172288A teaches a method for the continuous crystallization of 21 ' fructose from an aqueous solution. The syrup is rapidly mixed 22 with seed and put onto a surface until a cake is formed, which 23 is then conminuted to a free flowing granular product.
24 Although this method avoids the problem of continuous handling 2' of viscous solutions, the granular amorphous product contains all of the impurities that were in the feed syrup. In 27 addition, the extra grinding and drying stages raise the 28 operation costs considerably. Similar costs are incurred using 29 the method described in U.S. Patent No. 4,199,373, wherein syrup is seeded with crystalline fructose and allowed to stand _z_ in a mold or container, after which the crystalline material is recovered, dried, and ground.
Several patents describe processes wherein fructose is allowed to selectively crystallize from an S aqueous solution. In Japanese application 118,220, corresponding to the Japane Patent publication 60-118,200, published on June 25, 1985, two towers, one for Braining and one for crystallization are described. Feed from the first tower, containing 33-50% fructose syrup, is mixed with massecuite (crystal-containing) overflow from the second tower. The resultant mixture is cooled as the product moves downward in laminar flow. The crystalline fructose is then obtained by centrifugation.
Although this process avoids the additional drying and grinding steps of other crystallization proceses, its productivity is low and the scale up capacity is limited because of the necessity for vertical laminar flow and heat transfer demands.
One effective procedure for crystallization of fructose from aqueous solutions is described in U.S.

3,928,062. The patent described a method wherein a supersaturated solution is seeded and then evaporated and/or cooled under moderate stirring while maintaining the concentration and temperature within certain ranges.
By continuously concentrating the mother liquor, it can be used to produce multiple crops of fructose crystals.
Although a suggestion is also made that cooling alone can be used, such a procedure is not considered as advantageous as those using continuous evaporation because the mother liquid must be reconcentrated at the start of each batch. Although such a procedure is useful for producing small batches of crystalline fructose, such a process could not be used in an industrial scale production due to heat transfer constraints as well as lack of adequate mixing and control of 1 j supersaturation.
According to U.S. 3,°33,305, large fructose ciystals 3 are obtained in a two stage batch method from avatar solution by adjusting the pH of the solution and slowly cooling the mass to create a supersaturated solution which, when seeded, is optimal for crystal formation. Because of the long crystallization time of the process, a pH adjustment must be done and the productivity of the method is only about 0.25 t/m3/d.
Although all of the above processes have been used successfully for the production of crystalline fructose, it has 11 heretofore been thought to be impossible to produce crystalline 12 fructose on a large scale with high yields, high capacity 13 (productivity) and good urit from ita a p y queous solutions 1~ without resorting to costly processing steps including evaporating, drying, and grinding. An object of the present '-° invention is to provide a cost effective method for large scale, high capacity production of fructose crystals in high 18 yields.
19 Another object of the invention is to provide a method for crystallization of fructose which does not require 21 the use of organic solvents and without the need of pH
22 adjustment.

Still another object of the invention is to provide a 24 crystallizing apparatus that has optimal heat transfer and 2' mixing capacities for large scale production of high purity 26 fructose crystals.
Further objects will be evident from the description 28 of the invention which follows.
29 ' SUrL~SA.RY OF THE INVENTION
Disclosed herein is a method for producing I I
1 ~ crystalline anhydrous fructose whereby a small amount of crystalline fructose, providing a nucleation site, is added to 3 a fructose solution or crystalline seeds are allowed to form spontaneously in the solution. In a multistage crystallization process, all, stages except the first are seeded with a crystal foot, which is a mass of crystals and mother liquid (massecuite) from a previous crystallization. The resulting mixture is mixed while cooling slowly to carefully maintain the temperature and degree of saturation for anhydrous crystallization. , 11 Tn the production of fructose crystals, low 1z supersaturation and a small differential should be maintained.
13 In a preferred embodiment, the temperature differential between 14 the solution and the means used for cooling the solution is less than about 6°c, and the fructose solution, although 1° supersaturated, has a supersaturation of no more than 1.25, preferably between 1.1 and 1.2. Such conditions can most 18 readily be controlled in a heat transfer apparatus or 19 1 crystallizes whereby a heat transfer surface of at least 5/m2/m3 is provided. When such a crystallizes is used, it is 21 not inclined more than 45 degrees, and it contains means for 22 effective mixing, as well as cooling plates optimally spaced 23 not more than 300 mm apart and having an open sector in the 24 cooling plates of at least 5 degrees along the crystallizes.
2S In this embodiment, the mixer blades are located in '6 between and not more than 30 mm from the cooling surfaces.
Preferably, the velocity of the mixer blades is at least about 28 50 mm/sec during the crystallization process.

BRIEF DES CRTPTTON OF TFiE DRAWIDiGS
Fig, 1 is a side view, partially in cross section, of _5_ 1 ~ a fructose crystallizing apparatus according to the present I
invention.
3 Fig. 1A is a right side elevated view, partially in cross section, of a crystallizing apparatus.
Fig. 13 is an enlarged partial section view of the crystallizer shown in Fig, 1A.
Fig. 2 is a side view, partially in cross section, of another embodiment of the invention.
DETAILED DESCRIPTT_OPI OF THE INVENTION
A. Process in General i 11 According to the present invention it has now been 12 found that it is possible to improve.fructose crystallization ' 13 from water solution by a method where a horizontal cylindrical 1~ crystallizer is used both to allow efficient heat transfer 1~ within a small temperature differential and to effectuate good 1° mitring of the mass. Although it is not intended to be a 1~ limitation to the invention, it is believed that the parameters 18 described herein are adapted to create a dynamic equilibrium 19 between crystalline anhydrous fructose and dissolved fructose such that the growth of the crystalline structure is 21 sufficiently slow to avoid entrapment of water molecules.
22 The crystallization is carried out by seeding a 23 saturated or supersaturated fructose solution with a proper "'~ amount of seed crystals or allowing the solution to form seeds spontaneously, and then cooling the massecuite according to a 26 gradient which is determined during the crystallization. In a 2~ multistage crystallization rocess the f ..
p stages rom second ~.o 28 final crystallization are seeded with the proper amount of 29 crystalline foot. The proper amount of seeding crystals (Ms) depends on their size (ds), their length (m), on the quantity i ~:_:.
.. ~~Q~'~~~~~
;.
' i of the finished cr,rstals (M), and the desired crystal size (D) as follows:
3 Ms (tons) _ (ds/D)3 x M (tons) The fructose mass is simultaneously mixed to ensure optimum heat transfer and maximum crystal growth rate within the mass. The crystallizing process is a batch process, but it can be made semi-continuous by interconnection of several similar crystallizers. A two-stage method is advantageous if large crystal size of the product is preferable. The cooling l0 program depends on the quality of the feed syrup, but the 11 productivity is typically 0.5-0.8 t/m3/d and cooling time is 12 typically 15-30 hours by this improved method.
13 , A crystal yield of 65a of dry substance can be 14 reached in the end of the crystallization. The recovery and 1' drying of the crystals are made by conventional methods. If to the yield is very high, air bubbles can be mixed, at no more than 200, into the mass before the crystals are separated from 18 the mother liquor to reduce the viscosity. This makes the 1° centrifuging easier. The size of the product is typically 20 , 0.4-0.6 mm and the purity is over 99.5%.
21 B. Crystallizes 2 It is through the use of a crystallizes that the 23 conditions of supersaturation and optimal cooling, mixing, and 24 mass transfer can be accomplished in a large scale manner. For 25 large scale production of fructose, the crystallizes is 2° optimally about 30 m3 in size.
2~ With reference to the drawings, there are shown 28 crystallizers that are horizontal or inclined typically 5 29 degrees, but not more than 45 degrees, to ensure effective 30 axial mixing and drainage of the system. In a crystallizes, the heat transfer surface must be at least 5 mz/m3, preferably more than 10 mz/m3, so that the temperature difference between the fructose mass and cooling plates is not more than about 6.0°C, even if the cooling rate is 4°C/h.
With reference to Figures l, 1A and 1B, which depict an embodiement wherein multiple crystallizing zones are present, effective heat transfer is obtained when cooling water enters a cooling jacket 3 through an inlet 8 and circulates through cooling plates 2 which are situated inside the crystallizes and spaced not more than 300 apart. The cooling water passes through the cooling plates and out an outlet 9 located on the opposite end of the crystallizes from the water inlet.
A motor 6 mounted on a supporting stand 7 drives a shaft 4 which, at its pont of entrance into the crystallizing apparatus is surrounded by a sealing material 5. Strong mixer blades 1 extend from the shaft within the crystallizing apparatus. The crystallizes comprises mixer blades and cooling elements, which are situated inside the crystallizes. The mixer blades are situated between the cooling elements such that the distance between the blades and the cooling elements is not more than 30 mm. The mixing blades are situated between the cooling plates 2 so that the distance between the blades and the cooling plates is not more than 30 mm to ensure proper mixing of the mother liquid near the crystal surfaces. The rotation speed of the mixer is such that the velocity of the top of the mixer blades is typically 130 mm/sec but not less than 50 mm/sec at any moment of the crystallization. Small mixing efficiency was found to be insufficient to keep the crystal growth rate high while too much mixing resulted in spontaneous crystal forn~ation if supersaturation is high.
Fructose syrup to be crystallized (mother liquid) enters the crystallizes through inlet port 10. A horizontal flow in the crystallizes is effected by a small open sector in the cooling plates at least 5 degrees along the crystallizes.
_g_ 1 Massecuite containing solution is removed from the 2 crystallizing apparatus through outlet 11 whereupon it is 3 centrifuged to collect the crystalline material.
4 Referring to Figure 2, in another embodiment of 5 the invention, the crystallizing apparatus may contain 6 only two crystallizing zones. Such a crystallizes 7 employs the same general components of the crystallizes 8 shown in Figure 1, but effective heat transfer is 9 accomplished through circulation of cooling water through 10 a cooling water jacket 3' and into a single cooling plate 11 2' which extends upward through the center of the 12 apparatus. Similarly, only two mixing blades 1' are 13 necessary for mixing of the crystallizing mixture. The 14 motor 5' and shaft 4' are similar to the same components 15 in Figure 1.

16 C. The Crystallization Process 17 The temperature difference between the fructose 18 mass and cooling elements is kept less than about 6.0° C, 19 and the supersaturation is kept less than 1.25, 20 preferably between 1.1 and 1.2, during the whole 21 crystallization process. The sufficient heat transfer 22 area and mixing efficiency keeps the temperature 23 difference between the fructose mass and cooling plates 24 small enough despite very short crystallization times.

25 The supersaturation which determines the cooling rate is 26 calculated during the crystallization as follows:

27 Y - 10000 x (Ct-Cml) 28 Ct x (100-Cml) 3 0 Qml - 10 0 X F - Y

32 Cml - F (Qml, Tm) 34 S - Cml x (100 - Cml') 35 Cml' x (100 - Cml) 37 Y - crystal yield, % of dry substance amount 38 Ct - total dry substance concentration, % w/w 39 Cml - measured mother liquid concentration, % w/w _g_ i 2~~~t~~~
1 ~ Qml = mother liquid purity, ~ w/;o cf dry substance ' Cml' = saturation concentration of to mother liquid, ~ w/w F = experimentally measured solubility function Tm = temperature of the mass, °C
S = supersaturation The mother liquid concentration and temperature are measured by, for example, an on-line refractometer and a 6 suitable thermometer. The total dry substance concentration of the mass and the purity of the feed licuid are obtained from 8 laboratory analyses. The solubility cf fructose in water is a function of purity and temperature and is obtained experimentally.
11 The aqueous feed solution contains glucose as a'major 12 impurity, and it contains not less than 90~ by weight fructose 13 relative to the total weight of dry solids. The dry solids 1~ concentration of the mass must be hi her ~ ~-g than 90 w/w~ to ge~ a reasonable yield if the final temperature of cooling is about 16 25°C. The pH adjustment of the feed syrup is not necessary because of short crystallization times but the optimum pH range 18 of the feed syrup is 4.5-5.5 to minimize the degradation of.
19 fructose.
The careful supersaturation control., combined with 21 efficient heat transfer and effective mixing, results in '' maximal cr stal +
y growth rates wi,.hout spontaneous crystal 23 formation during the entire crystallization process. The productivity of 0.5-0.8 t/m3/d achieved in the main crystallizaticn by this improved method is substantially higher 26 than the yield obtained using the most advantageous method presented in the prior art.
28 In the preferred embodiment, the fructose solution is z9 placed in the crystallizer after being evaporated to a concentration of greater than about 90~ (w/w) dry solids and 1 adjusted to ~he saeding temperature. During this pre_ 2 crystallization phase, the seeding is mace ar.d tie cooling 3 program is determined as set forth above. Fellcwing this stage, a porticn of the mass is withdrawn, leavi.~.g a crystalline "fcot" which series as the seed in tae following main crystallization. Additional, concentrated feed is added and t::e cooli.~.g program continued once again as set out above.
After the main crystallization the crystals are separated from the ether li~sid by centrifugation and then dries.
In ar~cther embodiment, the crystal fect is used in 11 anoth~= crystallizes which is filled wit: additicnal syrup.

Crystallization F~rame'-~rs '- " '' r.
'' BOLT. t a pre- and main cr:stallization experi__~.,ents 1~ were done wi_: the 6 liter pilot crystallizes s:.cwn in F~c. ..
16 ecruipped with cooling water jacket and erfective mixer. The 1' crystallizes was connected with a prograa.:,able t :ermostat MQ-.~
Lauda RKP* 20. The length of the crystallizes was 18 c:~ and't:Ze diameter was 21-cm. The crystallizes has 42 m2/-~3 heat 20 transfer area, and it was slightly inclined.
21 The crystallizes consists of two crystallization 22 zones, the width of which were 8 cm, and two mixing blades were '3 installed in Seth zones. The distance be_-aeen the mixer blades 2'' and cooling plates was about 1.5 c:~. The rotation, speed o~ the mixer was 11 rpm and the velocity of the tzp of mister blades 2' was 130 ~nm; se= during the e:ca:~ples.

:~e feed s::rup was obtained fro::. an industrial plant 28 and =t COns~s=~~ Of ~~.5a fr'.lCtOS2, 1.0; deXtrOSe, 2.2°S
29 oligosacchar_des and the rest being zainiy salts as analyzed by gPLC. This s.~ru which had ocr ' p~ p c_-ystall~zing properties, was *Trade-mark _11_ ..
e~ ~ ~ ~~
1 chosen to demonstrate the effectiveness of the present ' invention. The pH of the feed syrup was 4.1, and it was 3 ~ adjusted to about 5.0 in all examples except No. 4.
'I The seed crystals were made from commercial fructose crystals by grinding with Fritsh pulverisette type 14.702. The 6 mean particle size of the seed crystals was about 0.03 mm and 7 900 of the crystals were between 0.02-0.08 mm as analyzed by a 8 PP!T-PAMAS particle measuring and analyzing system. The crystallization parameters of the examples are set forth in Table 1 and the results are listed in Table 2.

12 Table 1. ' 13 The Crystallization Parameters 14 Ex. 1 E:c. 2 Ex. 3 Ex. 4 Ex. 5 Ex. 6 1~ pre main pre main pre main pre main pre main pre main a 91.0 92.6 90.9 92 0 91.3 93 3 V 91.3 91.8 90.9 92 4 ~ 91 0.92.2 16 b 8.1 8.2 8.1 7.9 8.2 8.6 8.1 8.4 7.9 7.3 8.2 8.8 c .014 17.7 .038 14.2 .038 9.7 .038 10.8 .036 15.5 .038 10.3 17 d 56.5 57.0 56.0 57.0 57.0 57.0 56.0 56.5 ~ 56.0 57.0 56.0 57,0 a 35,5 28.0 36.0 28.0 40.0 28.0 37.0 24.5 35.0 28.0 37.0 25.0 18 g 24 21 26 24 24 15 24 20 72 24 24 30 19 = 5.0 5.1 5.0 4.1 5.0 4.9 21 a concentration of the mass, % w/w b amount of the mass, kg c amount of the seed crystals or crystal foot, % w/w ~4 of dry substance d initial temperature of the crystallization, °C
26 a final temperature of the crystallization, °C
27 f crystallization time, h 2$ g pH of. the feed syrup ~ In Example 1, the fructose solution was first l adjusted to pH 5.0 with 5~ w~'a NaHC03 solution. The f=_ed svru~
was evaporated to 91.0o w/w and 8.1 kg cf it was transfers=d to 3 the crystallizes which temperature was 56.5°C. When the 4 crystallize= ;aas fir ed, the syrup was seeded with 0.014, seeds and the cooling prcgsam of the pre-crystallization- Was started.
6 The concentration of the mother liquid was measured with a laboratory refractometer, and the mass Was cooled to 35.5°C so a that the calculated supersaturaticn maintained was less than i.25. The du=anon of the c~ystallization was 24 h, and the yield was 44.3% in the end of the crystallization.
11 when the pre-crystallization was finished, a part e?
12 the mass was pulled c=f and the rest was left in the 13 crystallizes se that the csystai feet, which determines the crystal size of the product, in the beginning of the main.
v5 CryStdl llZat'_O.~. Wa5 17. % o. Tile CyllSta1112er Was filled Wlth 15 eVapora~ed fe°_.~. Syrup WhlCh WaS miXed With the Crystal fOOt SC
that the temperature gradually rose to 57°C and the drv 18 substance concentration rose to 92.6%. The cooling proc~am was 1° started when the crystallizes was f'_lled. The mass was coclad 20 to 28°C so that the suaersaturation was r"aintained at less t:~an 21 1.25. The du=ation of the main crystallization was 21 h.
22 . After the main crystallization, the crystals were 23 separated from the mother liquid and washed by a laboratory 24 centrifugal Hettich Roto Silenta* 2. The diameter of the 2~ cents if~~ca 1 baske t was 21 cm and t::e amoun t of the washi.~,a '° water was 1.5-2.5o c~ the weight of h the mass. T.e crystals 2~ were dr_ed by a laboratory fluidizaticn d=-res.
2a The crysta'_ yield was 56.6% in the end of the csystali~zatior., and the purity of the cs=stair was 990 of the 30 dsv substance. The mean size of the product was 0.49 m.'n and *Trade-mark ~~z~~-~~'~~
' ~ the standard deviation from the mean size was 47% as measured 2 by a sieve analysis.
3 The crystallization procedures as set forth in the 4 remaining examples all had the same operation stages of EY.~,MPi:E
1. The variables were measured during the experiments as 6 described in Tables 2 to 7. The time from the beginning of the cocling in the main crystallization, the cooling water 8 temperature, the concentration and supersaturation of the i mother liquid are listed. In each case, the concentration was measured by a laboratory refractometer. The temperature 11 difference between the cooling water and the mass was less than i 12 1.0'C.

Examale 1 14 Table 2. Variables in Test Run T1 pre cryst. main cryst.
16 Time Temp Conc s Time Temp Conc s 7 h °C w/w % - h °C w/w 0.0 56.5 91.0 1.16 0.0 57.0 91.2 1.15 18 12.9 52.2 90.5 1.23 0.5 57.0 91.1 1.15 14.1 51.1 90.2 1.23 11.3 50.5 88.5 1.03 l~ ' 15.6 49.7 89.6 1.20 12.5 50.0 88.3 1.02 16.6 48.8 89.4 1.20 14.3 46.9 88.1 1.08 17.5 48.0 89.2 1.20 15.9 44.3 87.7 1.11 18.5 47.0 88.7 1.16 23.0 28.0 84.5 1.13 21. ' 20.3 42.9 88.3 1.23 24.9 35.5 84.9 1.05 ?~~-~~~t~'~
1 .I
::~:atole ,. 3. in Table Variables Test Run ~2 4 pre ca~'sW maincryst.
Ti me emp Conc s Time TempConc s h C w/w - h C w/w % -'-0.0 56.0 90.9 1.17 0.0 57.090.8 1 0.5 56.0 90.9 1.17 0.5 57.090.4 .

4.1 54.0 90.9 1,24 1.6 56.290.3 .

10.7 54.1 90.2 1.14 2.5 55,590.0 .

22.0 42.9 87.7 1.16 (42.7 28.083.2 .
8 1.02) 23.5 39.8 87.0 1.15 24.5 37.8 86.4 1.14 26.0 36.0 85.6 1.10 26.8 36.0 85 1 t0 .

al Example 13 Table 4. in Run n3 Variables Test 1.
Y

pre maincryst.
15 Ti cryst.

me Temp Conc s Time TampConc s h C w/w - h C w/w %

0.0 57.0 91.3 1.21 11.2 41.087.2 1 1' 12 3.6 56.5 91.1 1.18 12.5 37.086.0 .

18.0 49.3 88.9 1.12 13,0 35.485 .

19.2 48.0 88.9 1.15 14.6 30 . .

. . 1.10 19 20.1 46.9 88.5 1.14 15 28 . . 84.5 1.13 21.4 44.7 87.6 1.10 22.7 42.3 86.9 1.08 24.1 39.9 86.5 1.10 _15-~i I

I
1 i ~tam le ' 2 Table 5. Variablesin Test n4 Run 3 ' pre crys;.- main cryst.

4 Time Temp Conc s Time Temp Conc s h C w/w o - h C w/w -o 0.0 56,0 91.3 1.23 14.6 39.6 86 1 6 0.5 56.3 91.0 1.17 15.5 37.7 . .
86.2 1.12 3.2 55.8 90.0 1.19 16.5 35.5 85.7 1 7 20.5 45.1 87.4 1.07 17.5 32.8 85.2 .
1.12 21.9 42.4 86,9 1.01 18.5 29.9 85 1 8 23.2 39.7 86.6 1.11 19.5 27.0 . .
84.5 1.17 i 24.5 37.2 85.7 1.09 20.0 24.5 83 1 . .
I

i i Example 5 i Table 6. Variablesin Test ~5 Run pre cryst main cryst.

Time Temp Conc s Time Temp Conc s h C w/w ~ - h 'C w/w 16 0.0 56.0 90.9 1.17 0.0 57.0 91.1 1.15 5.5 55.5 90.9 1.18 1.3 56.2 90.5 1.09 70.6 36.1 85.3 1.07 16.8 43.0 87.3 1.09 18 19.0 39.5 86.5 1.09 21.9 33.3 85.1 1.09 24.0 28.9 84.4 1.11 _1 6, _ ~ ~
n Example 1 Table 6 3 7.
Variables in Test Run #6 pre ryst main cryst.
c 4 Time Temp Conc s Time Temp Concs h 'C w/w ~ - h C w/w %
-0.0 56.0 91.0 1.18 0.0 57.0 90 1 6 0.8 55.9 90.8 1.17 14.5 39.8 . .
87.01.14 l00~.8 47.1 87,9 1.07 15.4 38.0 86.01.08 22.1 4i.9 86.8 1.09 16.4 35.6 85.41.08 22.S 40.5 86.4 1.08 18.1 30.7 84 1 8 23.9 38.2 85.9 1.08 18.8 28.9 . .
84.31.11 24.5 37.0 85.7 1.09 20.0 25.3 83 1 9 . .

l0 11 Table 8.
Results of Test Runs in Examples 12 Ex. 1 Ex. 2 Ex. 3 Ex. 4 Ex. 5 Ex. 6 13 nre main pre main pre main pre main pre mainpre main a 44.3 56.6 40.5 57.5 38.8 62.8 42.6 58.840.9 57.8 4 54.5 41.9 '' b - - 60.6 56.2 58.?57.1 c 0.17 0.49 0.16 0.62 0.16 0.66 0.13 0.620.13 0.37 0.35 0.19 d 47 31 27 59 38 57 a 99 99.5 99.9 - 99.9 0.52 0,75 0.46 0.70 0.48 0.90 0.52 0.670.50 0.78 0.71 0.17 18 a crystal yield in the end of the pre 19 crystallization and before gingin the centrifu main crystallization, % w/w '1 b crystal yield of the product, % w/w c mean size of the product, mm 23 d standard deviation from the ze the mean si of product, %

2S a purity of the product, % substance w/w of dry 26 f productivity, t/m3/day '7 The mean size and the standard the deviation of pro duct were measured by the and in sieve analysis the end of the pre-crystallization by a laboratory mic roscope.

Claims (12)

1. A method for crystallizing anhydrous fructose from water comprising :
a) preparing an aqueous solution containing at least about 90 % of a dry substance containing fructose, the fructose content of the dry substance being at least about 90 % by weight;
b) seeding the aqueous solution at a temperature of 50 - 60° C to produce a liquid solution;
and c) cooling the resulting mass at a controlled rate and with continuous mixing such that supersaturation of the liquid solution with respect to saturated fructose is less than a ratio of about 1.25 and the temperature difference between the warmest and coolest portions of said solution is no more than about 6° C, wherein said cooling is accomplished in a crystallizer with a heat transfer surface of 5 m2/m3 or greater.

2. The method according to claim 1, wherein the resulting crystals are centrifuged and dried.

3. The method according to claim 2, wherein air bubbles are mixed into the mass at a volume no more than 20 % of the volume of the mass before centrifuging.

4. The method according to claim 1, wherein the supersaturation is between about 1.1 and 1.2.

5. The method according to claim 1 wherein said crystallizer is horizontal or inclined no more than about 45 degrees, and comprises mixer blades and cooling plates which are situated inside said crystallizer with no more than about 300 mm spaces between the plates and having an open sector in the cooling plates of at least about 5 degrees along the crystallizer.

6. The method according to claim 1, wherein said crystallizer comprises mixer blades and cooling elements, which are situated inside said crystallizer and which mixer blades are between cooling elements such that the distance between the blades and the cooling elements is not more than 30 mm.

7. The method according to claim 5, wherein the mixer blades have a rotation speed which is such that the velocity of the top of the mixer blades is not less than about 50 mm/sec at any point of the crystallization.

8. The method according to claim 5, wherein the heat transfer area of the crystallizer is 10 m2/m3.

9. A crystallizer, suitable for high capacity production of crystalline fructose from a solution containing same, said crystallizer being constructed in such a manner that when it is used, it is not inclined by more than 45 degrees, said crystallizer having a heat transfer area of over 5 m2/m3, an effective means for mixing, as well as cooling plates optimally spaced not more than 300 mm apart and having an open sector in the cooling plates at least 5 degrees along the crystallizer, such that the temperature difference between the warmest and the coolest portion of the solution is no greater than about 10° C.

10. The crystallizer according to claim 9,wherein said crystallizer is horizontal or inclined no more than about 45 degrees, and comprises a mixer and cooling plates situated inside the crystallizer, said cooling plates being spaced apart no more than 300 mm and having an open sector of. at least 5 degrees along the crystallizer.

11. The crystallizer according to claim 10, wherein said mixer comprises blades situated between said cooling plates such that the distance between the blades and the cooling plates is no greater than 30 mm.

12. The crystallizer according to claim 11, wherein the mixer has a rotation speed that is such that the velocity of said blades is not less than 50 mm/sec at any point of the crystallization.