The viscous clear liquid or “mucilage” extracted from fresh cut aloe leaves was the prime matter for this study. The Aloe vera plantation was located in the municipality of Umán, State of Yucatán, 35 km far from the city of Mérida, in eastern Mexico.

We followed the next steps to extract the aloe juice: we carefully washed the aloe leaves with soap in order to remove residual pollution and they were subsequently rinsed and immersed in a 200 ppm solution of Citricdin Plus for 20 mi- nutes. Then, we cut the leaves edges, 15 cm to the apex and 5 cm to the basis. Finally, the bitter exudate was drained by keeping the leaves in a vertical position for 30 minutes. Another 1 cm large portion of the leaf basis was removed after that, in order to discard the residual exudate. The outer green rind was removed immediately after, by cutting first the serrated margins of the leaf 1 cm to the edge and then peeling off the cuticle. The inner leaf gel juice was extracted then, by means of an Osterizer blender, cleared by passing it through a gauze filter and stored at 0˚C for further use.

For filtering the Aloe vera gel juice we passed it first through a Whatman® Grade 1 filter paper and secondly through a 0.22 microns pore size Millipore membrane. Both processes were achieved with the aid of a vacuum pump.

Nine acid hydrolysis and heat treatments were carried out combining three different Aloe vera gel juice (AGJ) % w/w concentrations with three other values for sulfuric acid % w/w concentration. The following values were chosen for AGJ concentrations: 1.5%, 2.5% and 3.5%. And for the sulfuric acid concentrations: 0.10%, 0.25% and 0.40% w/w. The nine hydrolysis processes were labeled T1 to T9 (Table 1). After the hydrolysis treatment, the assays were neutralized at pH 7 ± 0.34, using NaOH 0.08 N and each sample was brought up to a total volume of 50 ml and frozen. Besides, previously frozen samples were lyophilized. Dry matter obtained from each lyophilization was dissolved and homogenized in 2 ml of distilled water and was centrifuged at 12,000 rpm for 30 minutes at room temperature (24˚C).

Three solutions of AGJ were prepared at concentrations of 1.5%, 2.5% and 3.5%. Each one was brought to a total volume of 50 ml using distilled water.

Three treatments were prepared subsequently (labeled TC1, TC2 y TC3), with the following concentration distributions: 15 ml of AGJ 1.5% plus 15 ml of distilled water for TC1; 15 ml of AGJ 2.5% plus 15 ml of distilled water for TC2 and 15 ml of AGJ 3.5% plus 15 ml of distilled water for TC3. Once the samples homogenized, hydrolysis by a wet heat treatment was achieved at 1.2 lb_f/in² for 15 minutes. Pressure was then lowered to 0 lb_f/in² and flasks were immersed in an ice bath in order to stop the chemical reaction. Samples were frozen and lyophilized until further use.

Two commercial enzymes produced by Novozymes were used: Pectinex® AR and Viscozyme®.

Concerning Viscozyme®, it is worthy to mention that the furnisher specifications and parameters for its utilization were carefully followed.

Concerning Pectinex® AR, an enzyme substrate of 3 ml of AGJ with 3 ml of citrate buffer 0.05 M pH 5 (corresponding ratio 2:1) was deposed in test tubes and homogenized; pH was measured and the test tubes were then put in double boiler incubation at 40˚C ± 2˚C; once this temperature was reached, 4 µl of the Pectinex® AR enzyme were added and hydrolysis was stopped after a reaction time of 30 minutes by putting the test tubes in an ice bath for 10 minutes.

The Aloe vera gel juice obtained after filtration was then treated by the phenol- sulfuric acid method [9] to determine total carbohydrates in the samples.

Reducing sugars present in AGJ after each hydrolysis treatment were assayed using the 3,5-dinitrosalicylic acid (DNS) method [10] .

Merck glass TLC plate, silica gel coated with fluorescent indicator F₂₅₄, size 20 × 20 cm were used; they were cut into four pieces, size 8.5 × 10 cm, leaving 1 cm margin at the top and bottom edges, the later being used to mark the row where solution containing the sample was applied to the plate and the former, to mark the solvent front or limit to be reached by eluent after running. The spots of solution containing the aloe gel were applied on the row at a distance of 1.2 cm from one to another.

We used three 1000 ppm solutions of glucose, galactose and arabinose as reference patterns for TLC and the AGJ and AGJ hydrolyzed samples mentioned above.

An experimental 3² design was used to treat chemical hydrolysis, the AGJ and the sulfuric acid being the independent variables and the reducing sugars determination being the response variable.

The obtained values for the retention factor (Rf) in the reference patterns (glucose, galactose and arabinose) were averaged and compared with those reported by Zweig and Sherma [11] finding clear similarities.

Total and reducing sugars were quantified in the aloe extract obtained after filtration with the 0.22 µm pore size membrane. Among the total sugars present in AGJ (4.233 mg/ml of glucose), more than 50% are reducing sugars (3.088 mg/ml of glucose).

Concerning the identification of monosaccharides present in AGJ using TLC, the results can be appreciated in Figure 1.

The AGJ samples were compared to samples of paper filtered Aloe vera gel juice (PFJ) and it was found that four marks coincided, with an Rf value of 61.2 for two of them and Rf = 77.5 for the other two; similar retention factors suggest that the two samples could contain the same substance, the one obtained by filtration with a 0.22 μm pore size membrane (AGJ) and the paper filtered one (PFJ) [12] .

The results obtained from the nine treatments of AGJ by H₂SO₄ hydrolysis were analyzed under ANOVA setting and using the SPSS software (version 17, SPSS Inc.). The aim of this analysis was to evaluate the reducing sugars concentration (mg/ml of glucose) present in each treatment; the reducing sugars (RedSu) were taken as the dependent variable in the analysis. Furthermore, we also tried to discern the most significant treatments and to notice if there exists an interactive effect due to the combined action of the independent variables (acid and gel); this is caused by the 3² experimental design, as the three sulfuric acid concentration levels are confronted to the three AGJ concentration levels.

A comparison between the different values of mean square, by a Type III sum of squares, was achieved using SPSS software in order to determine if any significance exists between treatments. And, according to this analysis, a significance of 100% was found in every treatment.

Owing to the fact that all the treatments presented a significance of 100%, a descriptive analysis was carried out for each independent variable (acid or gel) with respect to the dependent variable reducing sugars (RedSu). Details of this analysis, based on Tukey’s method, can be observed in Figure 2, where treatments by acid hydrolysis are represented. The best results were obtained for those treatments corresponding to a 0.10% sulfuric acid concentration, i.e., the ones labeled T1, T4 and T7. The best among the three of them is T7, which corresponds to a volume of 15 ml of AGJ 3.5% solution and a 0.10% sulfuric acid concentration.

TLC results for the nine treatments are shown in Figure 3. Table 2 represents Rf values for the pattern samples and for the H₂SO₄ hydrolysis treatments.

The averaged value of Rf in reference patterns (glucose, galactose and arabinose) was calculated for each one of the three TLC plates where hydrolyzed samples were applied. This is due to the fact that it was not possible to compare nine hydrolyzed treatments to the reference patters in a unique chromatography. As a result, we have that the Rf value of hydrolyzed treatment labeled number 4 corresponds to galactose.

It was found that the best among the wet heat treatments of hydrolyzed AGJ was the one labeled TC3 (15 ml of AGJ 3.5% and 15 ml of H₂O added) for the am- ount of reducing sugars (mg/ml of glucose) under the pressure and time conditions described above (see Figure 4).

Comparing the Rf values on the glucose, galactose and arabinose reference patterns to the ones on the wet-heat-treated hydrolyzed aloe gel juice (TC1, TC2 and TC3), no matches were found (see Figure 5). Something similar was found when comparing the Rf values on reference patters to those on the non-treated Aloe vera gel juice samples (AGJ and PFJ). Nonetheless, if one compares the Rf

values obtained for the three wet-heat-treated samples (TC1, TC2 y TC3), it is observed that marks on TC1 and TC2 coincide with Rf = 59.4. However, there does not exist such a development for TC3.

It is obvious that in TLC results, where enzymatic treatments were applied, (see Figure 6) no monosaccharide separation is observed. Despite that, when comparing the results given for both enzymes, it is found that Pectinex® AR is the best for using on TLC.

Nevertheless, the results obtained in the quantification of reducing sugars for samples treated with enzymes are larger than for samples otherwise treated (by means of acid hydrolysis or wet heat). Moreover, there was no significant difference between both enzymes, as can be seen in Figure 7.

Figure 8 shows the compared results for the amount of reducing sugars (mg/ml of glucose) obtained by different hydrolysis treatments: the chemical one, with H₂SO₄, the one using wet heat at 1.2 lb_f/in² for 15 minutes and, finally, the enzymatic one. Comparison is made with respect to total sugars and reducing sugars in non-treated AGJ samples.

Regarding the chemical hydrolysis with H₂SO₄, the best treatment was the one labeled T7 (15 ml of AGJ 3.5% and 15 ml of a 0.10% H₂SO₄ solution), where 0.045 mg/ml of glucose were obtained together with a standard deviation of

±0.0017. For the wet heat treatments at 1.2 lb_f/in² for 15 minutes, TC3 (15 ml of AGJ 3.5% and 15 ml of H₂O) was the most efficient one, with 0.292 mg/ml of glucose and a standard deviation of ±0.0087.

In the case of enzymatic treatments, both of them showed high results. For the treatment using Viscozyme® (3 ml of AGJ with 3 ml of citrate buffer 0.05 M pH 5 and 4 µl of the Viscozyme® enzyme) 3.302 mg/ml of glucose were obtained and a standard deviation of ±0.0166. For the treatment with Pectinex® AR (3 ml of AGJ with 3 ml of citrate buffer 0.05 M pH 5 and 4 µl of the Pectinex® AR enzyme) 3.282 mg/ml of glucose were obtained together with a standard deviation of ±0.0327.

Among the total sugar (4.233 mg/ml of glucose) present in non-treated AGJ samples, 72.95% of them are reducing sugars (3.088 mg/ml of glucose).

From that 72.95% of reducing sugars one can obtain the 1.44% (0.0446 mg/ml of glucose) using treatment T7, involving an H₂SO₄ hydrolysis; the wet heat tre- atment (at conditions described above) produces 9.45% (0.292 mg/ml of glucose) of total reducing sugars, if one follows treatment TC3.

Concerning the enzymatic treatments, the non-treated solution (3 ml of AGJ with 3 ml of citrate buffer 0.05 M, pH 5 and corresponding ratio 2:1) produced 96.6% (2.986 mg/ml of glucose) of reducing sugars with a standard deviation of ±0.031; on the other hand, solution treated with Viscozyme® produced 106.9% (3.302 mg/ml of glucose) of total reducing sugars and solution treated with Pectinex® AR produced 106.3% (3.282 mg/ml of glucose) of total reducing sugars.

Regarding chromatography results, Measurements on TLC plates using a solution of non-treated Aloe vera gel juice gave the following values for the reference patterns: Rf = 50 for glucose, Rf = 46 for galactose and Rf = 55 for arabinose; none of these values matched those found for analyzed samples. However, two patterns coincide in their Rf values among the AGJ samples, one with Rf = 61.2 and the other one with Rf = 77.5; similar retention factors suggest that the two substances could be the same, one present in Aloe vera gel juice obtained by filtration with a 0.22 μm pore size membrane (AGJ) and the other one present in Aloe vera gel juice filtered with Whatman® Grade 1 filter paper (PFJ). TLC sho- wed a larger scanning for the PFJ sample than for the AGJ sample due to the removal of some solid substances after filtering it with the 0.22 µm size pore membrane.

Worth mentioning, the average value of Rf in reference patterns (glucose, galactose and arabinose) was calculated for each TLC plate where acid hydrolyzed samples were applied. This is due to the fact that it was not possible to compare nine hydrolyzed treatments to the reference patterns in a unique chromatography.

The Rf values for the reference patterns were the following: Rf = 55 for glucose, Rf = 50 for galactose and Rf = 57 for arabinose. The Rf values for the samples analyzed are listed right away: T1 showed Rf = 52.5; T2 an Rf of 47.5; T3 an Rf of 46.6; T4 an Rf of 50; T5 an Rf of 49.3; T6 an Rf of 52.2; T7 an Rf of 50.6; T8 an Rf of 50.7 and for T9, Rf = 52.8. According to the Rf values listed above, it can be said that galactose is present in treatment T4 and it would be equally present in T5, T7 and T8 (with Rf values of 49.3, 50.6 and 50.7, respectively).

TLC plate analysis for samples treated with wet heat showed the following Rf pattern values: 58.8 for glucose, 56.3 for galactose and 61.3 for arabinose. The Rf values for samples TC1 and TC2 were identical and equal to 59.4; as it can be seen, they do not match any of the pattern’s, and this means that this substance is none of the sugars taken as reference, but it can be concluded that the same substance is present in TC1 and TC2. The treatment labeled TC3 did not show any Rf value due to the fact that it was not developed in chromatography, as a result of the low concentration of the dry sample obtained after lyophilization. All the wet heat treatments needed a spot application on silica plates of over 50 µl and even though bands appeared very weakly.

TCL plate analysis for samples treated with enzymes did not show Rf values due to the lack of proper separation between compounds on the silica plate. This was caused by the wrong concentrations of enzyme and substrate in the enzyme-substrate complex used. It can also be explained by the characteristics of the enzymes used, as they were not specific for obtaining monosaccharides as glucose, galactose and arabinose.