SciELO - Scientific Electronic Library Online

Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Boletín de la Sociedad Chilena de Química

versión impresa ISSN 0366-1644

Bol. Soc. Chil. Quím. v.46 n.1 Concepción mar. 2001 


Roy Soto, Juanita Freer, Nazmy Reyes and Jaime Baeza*.

Renewable Resources Laboratory, Faculty of Chemical Sciences,
University of Concepción, Chile. Box 160-C, Concepción, Chile,
FAX: 56-41-247517. Email:
(Received: July 31, 2000 - Accepted: October 6, 2000)


This work evaluates the effect of the composition of the solvent mixture on the extraction of polyflavonoids from the bark of Pinus radiata. Non-reactive solvent systems were selected, specifically water 100%, methanol 100 %, and mixtures of them in proportions (7:3), (1:1) and (3:7). The amounts of the extracted with water were lower than those obtained with methanol mixtures. Large variations were not observed when the proportion of methanol in the solvent changes, however the reactive polyphenol percentage increases as the methanol proportion in the extracting solvent is increased. The effect of the section height of the tree from which the bark is obtained, on the nature of the polyflavonoids extracted with such solvents, has been evaluated. The obtained fractions were characterized by chemical assays, chromatographic (GPC) and spectrophotometric (IR) methods.

KEY WORDS : polyflavonoids, bark, Pinus radiata, tannins, polyphenols.


En el presente trabajo se evalúa el efecto de la composición de los disolventes metanol, agua y sus mezclas, sobre la extracción de poliflavonoides desde la corteza de Pinus radiata. Estas mezclas constituyen sistemas no reactivos. Los disolventes utilizados fueron específicamente agua 100%, metanol 100 %, y mezclas de ellos en proporciones (7:3), (1:1) y (3:7). Las cantidades de material extraído con agua fueron menores que aquellas con disolventes conteniendo metanol. No se observaron grandes variaciones en el rendimiento de la extracción al variar la proporción de metanol en el disolvente, sin embargo, el porcentaje de polifenoles reactivos aumenta conforme la proporción de metanol en el disolvente de extracción se incrementa. También fue analizado el efecto del origen de la muestra de corteza, con relación a la altura de la sección del árbol desde la cual fue obtenida, sobre la naturaleza de los poliflavonoides extraídos con tales solventes. Las fracciones obtenidas fueron caracterizadas por ensayos químicos, por métodos cromatográficos (GPC) y espectrofotométricos (IR).

PALABRAS CLAVES : poliflavonoides, corteza, Pinus radiata, taninos, polifenoles


Condensed tannins, a type of natural polyflavonoid, are present in many plant tissues, such as pine bark (Pinus sp.), acacia bark (wattle or mimosa), quebracho wood (Schinopsis sp.), etc.1-4

Since the wood industry in Chile constitutes an important economic activity5,6 related to the cultivation of Pinus radiata and Eucalyptus globulus, it is desirable to evaluate potentially more attractive and economically viable uses for by-products generated by this activity, for example, the pine bark that is currently rejected or underutilized.7

Many attempts have been made to optimize the isolation of these materials for different uses, and for that purpose different solvents have been utilized. These generally reactive solvents have included sodium sulfite8, sodium carbonate and sodium hydroxide solutions.9

However, this study is focused on obtaining a natural, non-modified tannin from Pinus radiata bark, in order to evaluate its reactivity and its application in adhesives formulations for wood. For this reason, it was of interest to evaluate a method for the extraction of the condensed tannins, varying the composition of the extracting solvent, and combining this variable, with the one which results from the original location of the bark used in the extraction, with respect to the height of the selected tree.

The solvents evaluated were water 100%, water:methanol (7:3), water:methanol (1:1), water:methanol (3:7) and methanol 100%. The extraction yield and some characteristics of the isolated tannins were determined. The other experimental conditions (T, solid/liquid ratio, etc.) were maintained constant.

In this way, the results obtained could guide future possible selection of bark with certain characteristics, for its utilization as raw material for the extraction of condensed tannins, or for its direct applications.


Bark of 27 years old Pinus radiata trees that grow in Concepción (Chile), was dried in a hot air current at an average temperature of 40-50 ºC. Once dried, the bark was classified as bulk bark, mean bark (smaller particles), bark with cambium and bark with other particulate material.

The bulk bark was ground to 40-60 mesh in a Retsch mill (model SM1) and extracted as described below. The moisture percentage of the ground bark was determined.

For the extractions, a modified procedure of that suggested by Ayla10 was used. Ten samples of ground bark of 25 g each were extracted in pairs, three consecutive times with 200 ml of the following solvents: water 100%, water:MeOH (7:3), water:MeOH (1:1), water:MeOH (3:7) and MeOH 100%. Each extraction was accomplished by heating for two hours in a water bath at 70ºC, while shaking with an agitation frequency of 100 cpm. After the first extraction, the hot mixture was quickly vacuum filtered, obtaining an extract and a solid residue. The latter was successively extracted twice by the same procedure. The extracts of the five pairs of samples, in each one of the extraction stages (3 for each pair), gave rise to 15 different extracts (Table I), which were concentrated, freeze-dried, weighed and analyzed as described below.

The Stiasny number for each freeze-dried extract was determined according to Suomi method.11

The samples (1,2)-1, (9,10)-1 and (5,6)-1, which represent the extremes and the middle point with respect to solvent composition, i.e. with water 100%, MeOH 100% and water:MeOH (1:1) respectively, were selected for study, including separation by Sephadex LH-20 and analysis by GPC and spectrophotometric methods.

Samples of the crude extracts were acetylated, according to García-Vallejo method12, and the acetylated products were analyzed by GPC. Another part of each one of these crude fractions (1 g) was filtered on Sephadex LH-20, following the procedures described by Makkar and Becker.13 The fractions eluted with acetone:water (1:1) were concentrated, freeze-dried, acetylated, and analyzed by GPC. Finally, another part of each one of these fractions (1 g) was purified by successive solid/liquid extraction (Soxhlet extraction) with ethyl ether (50 ml) and ethyl acetate (50 ml) for 4 hours, and the resulting products were then acetylated before analysis by GPC.

For evaluation of the effect of the height on tree, bark of a Pinus radiata tree, 25 years old, felled in Concepción (Chile), was sampled. The felled trunk was cut in three sections, each 4 meters long, corresponding to the lower, middle and upper parts of the specimen.

The cleaned bark was dried, ground and extracted, as has been described previously, with the variant that duplicate extractions with each one of the following solvents: water 100 %, water:MeOH (7:3), (1:1) and (3:7), and MeOH 100 % were carried out on bark samples of the three sections. This step gave rise to 30 different extracts, that were concentrated and freeze-dried separately.

The Stiasny numbers of each extract were determined, according to Suomi method.11

Another part of the extracts was acetylated, according to García-Vallejo et al.,12 and the acetylated products were analyzed by GPC, under the previously described operation conditions.

All GPC analyses were accomplished in a Merck-Hitachi chromatograph, model L-4250, provided with a Merck-Hitachi pump, model L-6200 A, and a Merck-Hitachi integrator, model D-2520. Three Styragel Waters columns serially connected (HR 0.5, HR 1 and HR 2, for molecular mass ranges of 0-1000, 100-5000 and 500-20000 Da, respectively) were used, at 60ºC, with a THF flow of 0.4 ml/min, and with UV detection at 280 nm. The columns were calibrated with polystyrene standards.


The results obtained from the three extractions stages with each one of the selected solvents are shown in Table I.

The greatest extraction is achieved in the first stage of each sequence (from 50% to 70% w/w of the total extracted), with medium yields in the second stage (from 12% to 25% w/w), and the least in the third stage (from 9% to 21% w/w). The amount of the extractives and the total percentage extracted increased with the methanol proportion in the extracting solvents mixtures.

The Stiasny numbers for the first stage of each one of the series analyzed with each solvent were generally greater than those obtained for the later stages. A trend of the Stiasny numbers with the composition of the solvent used in the extraction was not observed.

The reactive polyphenol percentage (calculated as the product of the extracted fraction by the Stiasny number) represents the w/w percentage of each extract able to react with formaldehyde on heating with acid. The value obtained for this parameter (Table I) is greater for the extracts of the first stage. In this stage, the percentage increases as the proportion of methanol in the extracting solvent increases. It is 11% w/w for water 100%, 24% w/w for solvents with 30% and 50% v/v of methanol, and 30% for solvents with 70% v/v and 100% of methanol. The same trend is observed for the total reactive polyphenol percentage (adding the yields obtained in the first two stages) with each one of the tested solvents. The smallest yield for the water 100% (15.81% w/w), intermediate yields for the water:methanol (7:3) and (1:1) systems (30% w/w approximately), and greatest yields for the water:methanol (3:7) and methanol 100% systems (35-36% w/w).

The identity of the extracted material as condensed tannin was confirmed by the IR spectra. All the extracts showed the characteristic bands for this type of material. The IR spectrum of the freeze-dried extract (1,2)-1, as an example, is shown in Fig. 1.

In Table II are shown the Stiasny numbers and recovery percentages of the fractions obtained by elution from Sephadex LH-20 with the solvents methanol:water (8:2), acetone:water (1:1) and acetone 100%, of each one of the extracts (1,2)-1, (5,6)-1 and (9,10)-1, in their crude form (non-pretreated), as well as after extraction with ethyl ether and ethyl acetate.

The low molecular mass polyflavonoids (non-tannins fraction) of the extracts were isolated by eluting the extracts through Sephadex LH-20 with methanol:water (8:2), and then the condensed tannins (tannins fraction) were eluted with acetone:water (1:1). The yields are summarized in Table II.

It was observed that for crude extracts passed by Sephadex LH-20, the non-tannin fractions represent 75.0% w/w, 70.9% w/w and 67.6% w/w of the extracts (1,2)-1, (5,6)-1 and (9,10)-1, respectively; while the tannin fractions represent 24.1% w/w, 24.8% w/w and 27.1% w/w, respectively. The Stiasny numbers for non-tannin fractions range from 64 to 82, being greater in the samples extracted with the methanol rich solvent. In the tannin fractions, the Stiasny numbers were between 93-95, values considerably greater than those obtained for the non-tannin fractions.

On the other hand, for the previously purified extracts, by successive extractions with ethyl ether and ethyl acetate, the non-tannin fractions represent 60.3% w/w, 59.4% w/w and 54.2% w/w of the extracts (1,2)-1, (5,6)-1 and (9,10)-1, respectively; while the tannin fractions were 34.8% w/w, 36.4% w/w and 38.6% w/w, respectively. Nevertheless, the trends were similar to those observed for the crude extract, the amounts of the non-tannin fractions were smaller and those of the tannin fractions greater. These last fractions presented Stiasny numbers between 95-96%.

In this way, it has been demonstrated that a previous purification of the extracts by extraction with ethyl ether and ethyl acetate gave better yields from the filtration by Sephadex LH-20. However, the benefit of including a such step in the purification process is not considerable, and therefore would not be necessary.

Tannin is a polymer with a wide range of molecular mass and when removed from bark, the values and distribution depend on the isolation procedure. GPC experiments were carried out to estimate the molecular mass (M) and the molecular weigh distribution (MWD) of the different tannin samples. In order to diminish the adsorption and association caused by hydroxyl groups, the acetylated derivatives were used. In addition to minimize the interactions, the acetylated tannins were dissolved in THF (tetrahydrofurane), which was also used as eluent. Column calibration was carried out by using polystyrene standards.

The SEC permeation chromatograms of tannins showed multimodal distributions. In Fig. 2 are shown the chromatograms for crude (a), and for the purified tannins (crude+Sephadex LH-20 and pre-extracted +Sephadex LH-20) (b, c). The two main peaks were not well resolved, but two maxims are clearly observed, one at 36 min approximately (Mn =27-37*103), and the other at 38 min approximately (Mn = 8-9*103). Besides this two peaks, in crude tannins appears another one at higher retention times, which is not present in the purified extracts, indicating that in the purification process the lower molecular mass portions were eliminated. The relative intensities of the two main peaks depend on the nature of the solvent used in the extraction. The greater the methanol greater is the proportion of the high molar mass fractions.

Related to the condensed tannins extraction from the bark of three height sections on a tree, Table III summarizes the results obtained with each one of the selected solvents.

The amounts extracted with water were lower than those obtained with methanol mixtures. However, large variations in the extraction percentages (w/w) were not observed when the methanol proportion in the solvent varied, even at the lowest proportion used (30 %). The amounts extracted are smaller for the lower section of the tree, but almost equal for the other two sections. Nevertheless, in terms of reactive polyphenol percentage (calculated as the product of the extracted fraction by the Stiasny number), it was observed that this value increased as the methanol proportion in the extracting solvents increased. The values for this parameter are greater for the tannins obtained from the middle section bark of the tree for all the solvents used.

The results of GPC tests on the acetylated derivatives of the fractions are summarized in Table IV.

In general we observed three principal peaks in GPC. One at 36 min approximately (Mn=21-33*103) for extractions performed on the three types of bark with MeOH 100% and water:MeOH mixtures (7:3), (1:1) and (3:7); another at 38 min approximately (Mn=5*103) for all extractions; and the last one at 46 min approximately (Mn=8*102) only for the extractions with MeOH 100% on the three types of bark.

For the 36 min peak, the intensity is greater for the extracts obtained with water:MeOH (1:1) and MeOH 100%, than to those obtained with water:MeOH (7:3) and (3:7). It is smaller for the bark extracts from the middle section of the tree, except for the MeOH 100% extract, which intensity is slightly greater than the one obtained from the upper section. In terms of variations of Mn with the extracting solvent composition, these values are wider for the extracts obtained with water:MeOH (7:3) and (3:7), and are closer for the extracts from water:MeOH (1:1) and MeOH 100%. Finally, the Mn variation is relatively small for the three sections extracts obtained with water:MeOH (1:1) and MeOH 100%, while showing a decreasing trend for the extracts from water:MeOH (7:3) when going from the lower to the upper section of the tree, and a less regular behavior with the extracts obtained with water:MeOH (3:7).

With respect to the 38 min peak, its intensity tends to decrease when the MeOH proportion in the extracting solvent increases, being greater for all the water:MeOH mixtures extracts from the middle section of the tree, though of smaller magnitude for the water:MeOH (1:1) extraction. These intensities are smaller for the extracts obtained with MeOH 100%, showing a slightly rising trend when going from the lower to the upper section of the tree. For this peak, Mn magnitude variations with the solvent composition, for the three types of extracted bark, show a general decreasing trend with increments in the MeOH concentration in the extracting solvent, reaching closer values for the MeOH 100% case. On the other hand, Mn variations with the tree section, for the different solvents, except for water:MeOH (3:7), show a maximum value for the middle section of the tree, with a smaller scattering in the MeOH 100% obtained extract.

Finally, for the last peak at 46 min, present only when extractions with MeOH 100% were carried out, their intensities, as well their corresponding Mn values, are maxima for the lower tree section and minima in the middle one.

Further studies are currently in progress on the use of these different extracts in resins synthesis.


The authors thank to Universidad de Concepción, FONDEF (grant D97I1055), German Academic Exchange Service (DAAD), Instituto Tecnológico de Costa Rica (ITCR), Universidad Nacional de Costa Rica (UNA) by financial and institutional support. This study is part of the R. S. Doctoral thesis.


1. E. Pulkkinen and S. Peltonen. Gel Filtration of Polyphenols from Scots Pine (Pinus sylvestris L.) Bark. Pap. Puu, 11, 687 (1980).        [ Links ]

2. A. Pizzi and A. Stephanou. Fast vs. Slow-reacting Non-modified Tannin Extracts for Exterior Particleboard Adhesives. Holz als Roh-und Werkstoff, 52, 218 (1994).         [ Links ]

3. E. Masson, A. Pizzi and M. Merlin. Comparative Kinetics of the Induced Radical Autocondensation of Poliflavonoid Tannins. II. Flavonoid Units Effects. J. Appl. Polym. Sci., 64, 243 (1997).        [ Links ]

4. S. Inohue, M. Asaga, T. Ogi and Y. Yasaki. Extraction of Polyflavonoids from Radiata Pine Bark Using Hot Compressed Water at Temperatures Higher than 100 °C. Holzforschung, 52, 139 (1998).        [ Links ]

5. T. Blackman. Exponential Growth Marks Chile’s Top Panel Producer. Wood Technology, 37 (1996).        [ Links ]

6. Directorio Forestal de Chile. (1997). 3th ed., Norton Ediciones, Chile.        [ Links ]

7. A. Pizzi. Tannin-Based Wood Adhesives. In Advanced Wood Adhesives Technology, 10th ed., Marcel Dekker, Inc., New York, pp. 149-218 (1994).        [ Links ]

8. E. von Leyser and A. Pizzi. The Formulation and Commercialization of Glulam Pine Tannin Adhesives in Chile. Holz als Roh-und Werkstoff, 48, 25 (1990).        [ Links ]

9. G. Vázquez, G. Antorrena and J. C. Parrajó. Studies on the Utilization of Pinus pinaster Bark. Part 2 : Kinetics and Yields of Alkaline Extractions. Wood Sci. Technol., 21, 155 (1987).        [ Links ]

10. C. Ayla. Wood Adhesives : Present and Future, John Wiley & Sons : New York, 69 (1983).        [ Links ]

11. L. Suomi. Bark Extracts and Their Use in Plywood Bonding. Pap. Puu, 2, 65 (1985).        [ Links ]

12. M. C. García-Vallejo, E. Cadahía and E. Conde. Hydrolizable and Condensed Tannins in Wood and Bark of Eucalyptus. Fifth Brazilian Symposium on the Chemistry of Lignins and other Wood Components, Proceedings, Vol VI, pp 176-185.         [ Links ]

13. H. P. S. Makkar and K. Becker. Isolation of Tannins from Leaves of Some Trees and Shrubs and Their Properties. J. Agric. Food Chem., 42, 731 (1994).        [ Links ]

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons