Colourants - Separation by GPC

By Bento, Luis San Miguel
Posted on 2008-07-09    Last edited on 2009-09-19

To separate natural colourants in sugar solutions, chromatography techniques have been used in the past. Godshall and Clarke (1988) used GPC (gel permeation chromatography), Sephacryl S-500 column, with a spectrophotometer detector associated in series with an RI detector (refraction index). In these tests sugar colourants were previously concentrated through dialysis.
Here is presented a separation of colourants and HMW compounds of beet and cane products using GPC. These compounds were detected using a Diode Array Detector (DAD) and an Evaporative Light Scattering (ELS) detector in series. Bento et al. used this technique to study high molecular weight (HMW) compounds in cane sugar products.
With this arrangement, one can detect compounds, as polysaccharides, not detected by spectrophotometry at 280 nm or higher wavelengths.
In this study, samples of beet and cane sugars are analysed and the results compared.

Evaporative Light Scattering Detector is a sensitive detector for non volatile solutes in a volatile liquid stream. Eluent enters the detector at the top of the evaporation chamber (Fig. 1).


Figure 1 - ELS detector

The eluent/solute is fed into the nebulizer assembly where an air operated venturi jet atomises the eluent into an uniform dispersion of droplets which then pass as a continuous stream into the evaporator. When pure eluent is being evaporated, only its vapour passes through the light path and the amount of light scattered to the photodetector is small and gives a constant response. When non-volatile solute is present, a particle cloud passes through the light, causing light to be scattered. This scattered light enters the optical aperture of the detection system and generates a signal response from the photomultiplier. The signal is amplified and gives a voltage output results. The quantity of light detected is dependent on the solute concentration and solute particle size distribution (PL-EMD960 Operator’s Manual).


Pharmacia FPLC system equipped with a liquid Chromatography Controller LLC-500 Plus and High Precision Pump P-500.
Column: Pharmacia, 300 mm length and 10 mm internal diameter, packed with Superose 12, cross-linked, agarose-based medium, with an average particle size of 10-11mm, and an exclusion limit of circa 2x106 g/mol (globular proteins).
Detectors: Lichograph Diode Array Detection System (DAD) L-4500 Merck-Hitachi, continuous absorption measurement in the UV/VIS range.
Evaporative light scattering detector (ELS) PL-EMD 960 Polymer Laboratories.
Laboratory Centrifugal: MSE, Basket 300.
Ultra-sounds: Bandelin, Sonorex TK30.
Membranes Nylaflo, Gelman Sciences.

Acetonitrile gradient grade for chromatography - MERCK.
Ammonium Acetate - MERCK.

Eluent and sample preparation

An exigency of the use of ELS detector is to use an eluent with low boiling point to be evaporated inside the detector. In this study we used distilled water with 30% acetonitrile, containing 0.005 M ammonium acetate. The eluent, at pH 7.0, was filtered through a 0.45 micrometer membrane and then degassed during 15 minutes with ultra-sounds.
Samples were diluted in eluent and adjusted to pH 7.0. Solutions were filtered through a 0.45 micrometer membrane and then degassed during 5 minutes with ultra-sounds.

Run method
Before each run, the system was stabilised for a period of 30 min with eluent at 0.30 ml/min. Run time of 80 min started with sample (200 microL) injection. Eluent flow rate was 0.30 ml/min. Effluent was analysed in DAD and ELS detectors placed in series.
Measurements with DAD detector ranged from 240 to 450 nm.
Conditions at the ELS detector were: air flow at 5 l/min and air temperature 80 °C. For security reasons a flow of nitrogen was fed continuously into the detector case.

White sugar wash
White sugar, 50 g, calibrated between sieves 0.300-0.425 mm, was mixed with 50 g of methanol saturated with sugar. This mixture was introduced into a flask and placed in a water bath with a rotating system for a period of 30 min at room temperature. The mixture was then centrifuged at 4000 r.p.m. for a period of 5 min to obtain the washed sugar.

The tests described here used a Superose 12 column as described in Run Method. In order to protect the column, samples were filtered through 0.20 mm. With this filtration, compounds of very high molecular weight will be removed and are not considered in this study.
During gel chromatography, compounds of very high molecular weight (higher than 2000 kD with the column used) will pass the column with the velocity of the eluent forming the first group to be detected. Afterwards, colourants are separated by decreasing molecular weights. Compounds with molecular weights lower than the inferior separation limit of the column will be eluted in one final group. With Superose 12 this means circa of 1 kD. Sucrose, simple sugars and other small molecules will be eluted together at the end of the run. As we do not separate sucrose in the samples we can not detect in the ELS detector colourants of low MW as they are eluted mixed with sugars. Therefore, in this study, we are limited to compounds with MW higher than 1 kD that will be referred to as HMW material.
Due to the physical arrangement of detectors there is a time difference of 0.39 minutes between the two detections. In the following text, when not specified, retention times are referred to the ELS detector. In Figures, Rt are referred to the respective detector. Chromatographic column was calibrated in a previous work (Bento et al.). DAD results are presented at 330 nm. For simplicity, the expression “colour” means absorbancy at this wavelength.
In this study we test one cane raw sugar from Cuba and a beet raw sugar from Germany. Observing the results of cane raw sugar with ELS detector (Fig. 4), we can consider three groups of compounds. The first group (A), corresponds to retention times (Rt) between 20 and 30 minutes. This group comprises compounds of very high molecular weight, with more than 250 kD. As observed at Figure 2, this group of colourants has low colour intensity. These compounds must correspond to colourants associated to polysaccharides, as described by Godshall et al. (1992).


Figure 2 - Cane Raw Sugar - DAD detector


Figure 3 - Beet Raw Sugar - DAD detector

In a previous study with a cane raw sugar, Bento et al., observed that these compounds remain preferentially in sugar crystals after affination.
As observed at Fig. 3 and 5, beet raw sugar has a small quantity of these compounds.


Figure 4 - Cane Raw Sugar - ELS detector

At Table 1 is presented the percentages of peak areas, calculated on ELS detector values, of the three groups considered.
As observed, Group A represents 2.75% of HMW compounds from cane raw sugar sample and only 0.15% on beet raw sugar sample.

Table 1

Rt (min)
20 - 30
30 - 48
48 - 57
Molecular weight
> 250 kD
12 kD - 250 kD
2,5 kD - 12 kD
Raw cane sugar
Raw beet sugar
Cane white sugar
" (washed)
Beet white sugar
" (washed)

The second group (B), with Rt between 30 and 48 minutes, comprises compounds with high molecular weight and a high colour intensity. Molecular weight range from 12 kD to 250 kD. In this group we can consider two peaks with 38 and 40 min of Rt (B1 and B2) that presents high colour intensity (Fig. 5). In group B, other peak (B3) is observed with a Rt of 43 min. This peak must comprise compounds with low or no colour as it is not observed an equivalent peak in DAD chromatogram.


Figure 5 - Beet Raw Sugar - ELS detector

The Group B represents the majority of HMW compounds from the cane raw sugar sample, that is 56.69% of these compounds. For the beet raw sugar this value is 30.39%.
The next group (C) comprise compounds with lower molecular weights, between 2.5 kD and 12 kD, with Rt from 48 to 57 minutes. In this group we observe two peaks: C1 with 49 min and C2 with 51 min of Rt. This group represents 40.56% of HMW compounds in cane and 68.92% in beet raw sugar. This means that beet raw sugar contains half of the quantity of compounds with molecular weight higher than 12 kD present in cane raw sugar.
In the cane raw sugar sample it is observed that in Group C must exist a small quantity of low MW compounds with high colour intensity as can be observed comparing DAD and ELS results. This is not observed in the beet raw sugar sample.
Other tests were made using samples of cane white sugar with a colour of 57 IU and beet white sugar of 106 IU colour. Sugar was washed with methanol as described earlier. With this wash, colourants in the outside layer of the crystal were removed. Tests were made with original sugar and washed sugar. Results with ELS detector are presented at Figures 6 to 9 and at Table 1.


Figure 6 - Cane White Sugar - ELS detector


Figure 7 - Cane White Sugar (washed) - ELS detector

It is observed that two main groups of colourants remain in cane white sugar: Group A and Group C1. A small presence of Group B and C2 is observed. After methanol wash, Group C was reduced, from 41.27% to 27.41%, indicating that these compounds are preferentially in crystals syrup layer. Group A compounds of very high molecular weight remained inside the sugar crystals. These compounds remain through the refining process (Carbonatation + Resins) from raw to white sugar.
In the beet white sugar sample, compounds of group A are present in a small quantity, 0.78%. Group B represents 38.61% of HMW compounds, much more than in cane white sugar, that is, 3.64%.
It is observed also that the majority of HMW compounds in beet white sugar have a MW lower than 2.5 kD (60.61%) and this percentage decreases only slightly after methanol wash.


Figure 8 - Beet White Sugar - ELS detector


Figure 9 - Beet White Sugar (washed) - ELS detector

It is important to know the distribution of sugar colourants by their molecular weight and colour intensity. The technique described here will give this answer in a simple and rapid test with the advantage of using a not very expensive equipment. This technique uses medium pressure pumps, a GPC chromatographic column, an UV spectrophotometric detector and an Evaporative Light Scattering detector. The main advantage of this technique is to be able to analyse materials that are not detected by UV spectrophotometers. This technique can be important to study how decolourizers or separation systems, as membranes, remove or separate different colourant groups. We can also identify colourants more harmful to the refining process, that is, those that have a more tendency to be include in white sugar crystals.
In the study presented here it was observed that cane sugars have a more percentage of high molecular weight compounds that beet sugars. Also, very high molecular weight material, possibly associated with polysaccharides, present in cane sugars, are present only in a small quantity in beet sugars.
ELS detector can be associated with a more accurate chromatographic technique as HPLC or IC. By the results obtained here we can preview that Evaporative Light Scattering detector can give useful information to sugar related colourant materials in sugar factories and refineries.

Bento L.S.M., Pereira M.E., Sá S.,1997, Gel permeation chromatography of sugar materials using
          spectrophotometric and evaporative light scattering detectors, Proc. of SIT Conf., 383-392
Godshall M.A., Clarke M.A., 1988, High molecular weight (HMW) color in raw and refined sugars, Proc. of
          S.I.T. Conf.,
pp. 180-193
Godshall M.A., Clarke M.A., Miranda X.A., Blanco R.S., 1992, Comparison of refinery decolourization
          systems, Proc. of the Sugar Proc Res. Conf., pp. 281-305
PL-EMD960 Operator’s Manual for Light Scattering Detector, Polymer Laboratories (UK)


Comment on this entry
Creative Commons License 493367 visits since 2007-07-10 Privacy policy

Ler esta entrada em português