[Home] [APPW 2004] [Journal papers]

 

DEVELOPMENT OF AN ALTERNATIVE SOLVENT TO REPLACE
BENZENE IN THE DETERMINATION OF ORGANIC SOLUBLE EXTRACTIVES IN WOOD

Authors

Nelson L. Sefara and Mike Birkett

Company

Sappi Forest Products

Keywords

extractives, Tappi T204 test method, benzene-ethanol, wood

 

 

 

ABSTRACT

The determination of the amount of extractives in wood is commonly carried out by extraction with organic solvents using a Soxhlet apparatus in accordance with Tappi method T204. A solvent mixture consisting of benzene and ethanol (2:1 v/v) is widely recognised as the standard method for the removal of most extractives. However, benzene is a known carcinogen and it has become necessary to find a replacement solvent. In this study, a number of solvents (dichloromethane, toluene, toluene-ethanol, acetone and ethanol) were evaluated on sawdust samples of E. grandis and P. patula in order to determine a suitable replacement solvent for this analysis. The results showed that extraction of E. grandis sample with toluene-ethanol (1:1 v/v) and acetone yielded results that closely match those obtained from benzene-ethanol. Acetone offered better properties than toluene-ethanol and was therefore identified as a better replacement to benzene-ethanol

INTRODUCTION

The term, wood extractives, refers to a group of low molecular weight organic compounds that are extractable from wood by means of polar or non-polar solvents. The compounds comprise of fats, waxes, fatty acids, steryl esters, sterols, terpenoids, and other phenolic compounds,. Their composition in wood varies between 3-6% of the weight of the wood depending on the species. Generally, softwoods have higher extractives than hardwoods. Concentrations of extractives can be higher in certain parts of the tree such as branches, roots and damaged areas.

Wood extractives are known to have a negative effect on the pulp and paper making processes. Although the majority of the extractive compounds dissolve in cooking liquors during pulping, some are carried over to the bleaching processes and accumulate to form sticky deposits (so-called pitch deposit) on pulp that ends up in the paper making process. Pitch deposit on paper has a negative impact on paper strength properties because the resinous extractives tend to block the reactive groups on the surface of the fibres thus hindering inter-fibre bonding.  

The determination of wood extractives content is commonly carried out using Soxhlet extractors, in accordance with standard Tappi T204 om-88 test method. Solvents such as benzene, acetone,, ethanol, toluene, hexane, dichloromethane, methanol, diethyl ether, toluene-ethanol, etc. have been used.  Numerous publications have shown that extraction yields and composition of the extracted compounds differ depending on the solvent used.  In most cases, a combination of polar and non-polar solvents is used to maximize the extraction yields from wood. A solvent mixture consisting of benzene and ethanol (2:1 v/v) has been widely used by many laboratories because it removes most of the extractives and the results are highly reproducible between the different laboratories. Despite this advantage , many laboratories have been using alternative solvents because benzene is a hazardous carcinogenic solvent. Our laboratory has been using benzene-ethanol solvent and has over the years developed a database of extractive content for a range of species currently used at Sappi. It has become necessary to find an alternative solvent that will produce results which are comparable with those obtained from benzene-ethanol.

In this project, we compare the extraction yields and the repeatability achieved by a number of solvents on two groundwood samples of E. grandis and P. patula. The solvent that were evaluated include acetone, toluene, ethanol, dichloromethane and toluene-ethanol (evaluated at three different ratios). The results from this study were compared to those obtained using benzene-ethanol. The aim was to identify a suitable a solvent system with low toxicity levels, capable or removing approximately the same amount of extractives as those removed by benzene-ethanol and also has acceptable levels of precision. 

MATERIALS AND METHODS

A single tree of Pinus patula and of Eucalyptus grandis were randomly sampled from Sappi's Hodgson plantation in the Kwazulu-Natal region.  From each tree, a 1.5-m billet was cut at 20-cm from the base of the tree. The billets were transported to Sappi Forestry Research, where the billets were further cut into discs at every 150-cm intervals. The sawdust generated during the cross-cutting was collected to form a composite sample for each species.

The sawdust samples were transported to the Technology Centre where they were dried, milled and screened through 40-60 mesh (425 250 mm aperture) screens. The ground wood sample retained by the 250 mm screen was used for the analysis. Prior to extraction, the moisture content was determined in accordance with TAPPI method T210. Analysis of extractives was carried out in accordance with TAPPI method T204 om-88. The sample was extracted with benzene-ethanol (2:1v/v), toluene, ethanol, acetone, dichloromethane and toluene-ethanol. Toluene-ethanol mixture was evaluated at three different compositions, viz. 25%, 50% and 75% toluene content. All solvents were purchased from Merck and were of GR grade.

The extraction apparatus consisted of Soxhlet extractions units, which are connected to 250-ml flat-bottomed flasks and condensers. Approximately 5-g per sample was weighed into tarred cellulose thimbles and extracted with 200-ml of solvent for 4 hours at the rate of 4 - 6 extractions per hour.  Six replicate measurements were carried out for each solvent mixture. After extraction, the solvents was evaporated and collected in a condenser. The remaining extracts were dried in an oven overnight, weighed and expressed as percentage oven dried mass of the sample in the thimble. An analysis of variance (ANOVA) and a Duncan test was carried out to test for significant difference between the extracts of the solvents. All the results were tested at 0.05% level of error. Using the result from the replicate measurements, a repeatability ratio was calculated in order to obtain an estimate of the precision of each solvent.

RESULTS AND DISCUSSION

Effect of solvent on yield of organic soluble extractives

The physical and chemical properties of the solvents that were used in this study are summarised in Table 1. Acetone was the more polar compound while toluene was the less polar solvent. The toxicity data showed that benzene was the most toxic solvent while ethanol was the least toxic. Dichloromethane was also more toxic and toluene and acetone relatively less toxic.

Table 1: Physical and chemical properties of the solvent

Solvent

Polarity index, P`

Dielectric constant (at 25 C)

Boiling point, C

Toxicity information TWAa (ppm v/v)

Benzene

3.1

4.2

80

5

Toluene

2.4

2.4

110

100

Ethanol

4.3

24.5

78

1000

Acetone

5.1

20.7

56

750

Dichloromethane

3.1

8.9

40

50


aTWA time weighted average, a measure of the permissible exposure levels.

Figures 1 and 2 display the mean, standard error and range of the extracted removed by the different solvents from E. grandis and P. patula sample, respectively. It is evident from the two graphs that different solvents extract varying amounts of extractive compounds. The ANOVA and Duncan test revealed that the amount removed were significantly different for the different solvents that were used.

Figure 1

Figure 1: Organic soluble extractives from E. grandis. Identical letters illustrates that no significant differences was obtained.  BE = benzene-ethanol, ACT = acetone, DCM = dichloromethane, TOL = toluene, T75_E25 = 75% Toluene- 25% Ethanol, T50_E50 = 50% Toluene- 50% Ethanol, T25_E75 = 25%Toluene- 75%Ethanol and ETHNL = ethanol

Figure 2

Figure 2: Organic soluble extractives from P. patula. Identical letters illustrates that no significant differences was obtained. BE = benzene-ethanol, ACT = acetone, DCM = dichloromethane, TOL = toluene, T75_E25 = 75% Toluene- 25% Ethanol, T50_E50 = 50% Toluene- 50% Ethanol, T25_E75 = 25%Toluene- 75%Ethanol and ETHNL = ethanol

 

It is observed from Figures 1 and 2 that toluene and dichloromethane removed the lowest amount of extractives. Extractives removed by toluene amounted to 0.27% in E. grandis and 1.20% in P. patula. Extraction of the P. patula sample with dichloromethane removed significantly highly amounts of compounds than toluene, while no significant differences in the amount of extractives removed by DCM and toluene was obtained for the E. grandis sample.  The amount of extractives removed by toluene and dichloromethane were on average 40% lower than that removed by benzene-ethanol. The low extraction power of these solvents is attributed to the lower polarity index of the solvents and therefore lacking the ability to extract more polar compounds in the wood.

Using the conventional benzene-ethanol mixture, the extractives content in E. grandis and P . patula was found to be 1.89% and 2.06%, respectively. Extraction yields achieved by acetone were significantly higher than that removed by toluene and by dichloromethane. In E. grandis sample, acetone extracted significantly higher amounts than benzene-ethanol while a significantly lower amount was obtained for P. patula sample.

It is noticed from the figures that when toluene is used in combination with ethanol, the extraction yields increased. The results from E. grandis showed a linear increase in the extraction yields as the composition of the ethanol in the toluene-ethanol mixture was increased. P. patula also showed a slight increasing trend, reaching a maximum at 50%Toluene-50%Ethanol solvent mixture, followed by a decreasing trend in yield with further increases in the ethanol composition. The extraction of E. grandis sawdust with ethanol alone gave extraction yield of 5.25%, which were almost 3 times those removed by benzene -ethanol.  After inspection of the extracts, there was evidence of the presence of other non-resinous compounds in the extracts (e.g. short polymers and lignans), suggesting that ethanol by itself was not selective to extractives alone but also removed other low molecular weight wood compounds.

Using the extractive yields achieved from E. grandis and P. patula, the repeatability of each solvent was calculated in accordance with Tappi 1206 method and the results were plotted in Figure 3. In the figure, a low repeatability ratio illustrates that a high precision in the measurement. It is noticed, for both E. grandis and P. patula, that the 50%Toluene-50%Ethanol solvent gave the highest precision while toluene and dichloromethane had the lowest repeatability. The repeatability ratio obtained using benzene-ethanol, as a solvent was relatively poor in P. patula but average in E. grandis.

Figure 3

Figure 3: Repeatability ratio calculated for each solvent on P. patula and E. grandis extractives

 

Using the results from the yield and the repeatability of each solvent, a ranking order was established in order identify a solvent which exhibited properties that closely match those achieved by benzene-ethanol. The rankings (in increasing order) are summarised in Table 2.

Table 2: Ranking1 order for extractive yields for P. patula and E. grandis.

Solvent

E. grandis

P. patula

 

%Extractives

Repeatability

%Extractives

Repeatability

Benzene-Ethanol (2:1)

4

4

5

3

Acetone

5

5

4

4

Dichloromethane (DCM)

1

3

2

2

Toluene

2

2

1

1

Toluene (75%): Ethanol (25%)

3

1

7

6

Toluene (50%): Ethanol (50%)

5

8

8

8

Toluene (25%): Ethanol (75%)

7

7

6

5

Ethanol

8

6

3

7

1. Ranking in increasing order: i.e. the lowest number illustrates lowest percent extractives and worst repeatability

The 50%Toluene-50%Ethanol solvent mixture ranked very well in terms of repeatability and the amount removed from P. patula. Although, it also exhibited the highest repeatability in the E. grandis, it ranked average with regard to its ability to remove the majority of the compounds. The 75% Toluene-25% Ethanol mixture ranked well in terms of repeatability and yield on P. patula sample, but poorly on E. grandis. This is contrary to a recent study by Pepetini which revealed that toluene-ethanol (2:1) gave more reliable results and had the highest repeatability in comparison to benzene-ethanol. Ethanol (100%) also showed high levels of repeatability. The overall performance of benzene-ethanol mixture was average in both E. grandis and P. patula. Acetone extracted approximately the same amount of material as benzene-ethanol. This is evident in both P. patula and E. grandis. The repeatability values obtained for acetone compared very well with those achieved by benzene-ethanol. 

CONCLUSIONS

Based on the ranking results in Table 2, it would appear that the toluene-ethanol mixture (1:1) would offer good repeatability and also removed the maximum amounts of extractives. However, yield from this solvent was consistently much higher than that removed by benzene-ethanol. Solvents that gave yields that were comparable to benzene-ethanol were acetone and 75%Toluene-25%Ethanol. It is envisaged that if either of the two solvents is selected as a replacement solvent that the results would correlate very well with historical data obtained with benzene-ethanol. However, acetone offers better advantages than toluene in terms of the toxicity data. In addition, the toluene-ethanol mixture has been reported to have poorer reproducibility because it doesn't form a constant boiling mixture during the extraction period. This is attributed to the differences in the boiling points of the two solvents (Table 1). Acetone is therefore a preferred solvent and it will probably be adopted at Sappi as a replacement solvent to benzene-ethanol.

LITERATURE CITED

1. Fengel, D, Wegener, G., Wood: chemistry, ultrastructure, reactions, New York: W. de Gruyter, 1984

2. Sjostrom, E. Wood Chemistry Fundamentals and Applications, Academic Press, New York, 1981

3. Back, E. L., Allen, L. H. (Eds), Pitch Control, Wood resin and deresination, Tappi Press, Atlanta, 2000

4. Erkman, R., Eckerman, C., Holmbom, B, Studies on the behaviour of extractives in mechanical pulp suspension, Nordic Pulp Paper Res. J. 5(2), 96, (1990).

5. Tappi test method T204 om-88, Solvent extractives of wood and pulp, 1996 -1997.

6. Allen, L. H., Sithole, B. B., MacLeod, J. M., Lapointe, C.L., McPhee, F.J., The Importance of Seasoning and Barking in the Kraft Pulping of Aspen, J. Pulp and Pap. Sci. 17(3), 1991, pp J85 J91.

7. Dermis, A., Analysis of beech wood fatty acids by supercritical acetone extraction, Wood Sci. Tech. 25 1991, pp365-370

8. Sun, R. C., Tompkinson, J, Comparative study of organic solvent and water-soluble lipophilic extractives from wheat straw I: yield and chemical composition, J. Wood Sci.

9. Sunderg, K.; Pettersson, C., Eckerman, C., Holmbom, B., Preparation and Properties of a model dispersion of colloidal wood resin from Norway Spruce, J. Pulp and Paper Science, 22(7); pp J248 J252

10. Ekman, R, Holmbom, B., Analysis by Gas Chromatography of the wood extractives in pulp and water samples from Mechanical Pulping of Spruce' Nordic. Pulp Paper Res. J. 4(1), 1989 pp 16-24

11. Sithole, B. B., A rapid spectrophotometric procedure for the determination of total resin and fatty acids in pulp and paper matrices,  Tappi J., 76(10), 1993, pp123 -127

12. Mutton, D. B., Tappi (41(11), 1958, pp 632.

13. Caperos Sierra, A.; Romero Salvador, A. R., Garcia-Ochoa Soria, F., Kinetics of wood extraction with solvents, Tappi J., 74(5), 191, 1991.

14. Lewis, RJ, SAX'S Dangerous Properties of Industrial Materials, 10th Edition, John Wiley & Sons, Inc., 1999

15. Pepetini, F., Development of a method to quantify extractives in wood, MSc Thesis, University of Natal, 2003

BACK TO TOP