Printer Friendly

Production of furfural from cotton stalks.

Abstract

A one stage process was used to produce furfural from cotton stalks by hydrolysis in hydrochloric acid. A full factorial design of experiments was used to examine the effect of reaction time, liquor to solid ratio as well as the acid concentration (at constant temperature of 140[degrees]C) on the yield of furfural. A regression relating the effect of each parameter as well as binary and ternary interactions was deduced. Insignificant terms were deleted using the student test, and a regression having only four terms was left to express the effect of time, concentration and their interaction and a numerical constant on the yield, the adequacy of this regression to express the system was tested using Fisher's test, where the regression proved to be adequate. The optimum conditions for furfural production were obtained using the steepest ascent technique. These conditions are: as follows; the time of hydrolysis is 3.5 hrs, acid concentration 11.5% (w/s) and Liquid/Solid ratio = 30 : 1. The effect of adding some metal oxides such as calcium oxide, zinc oxide, aluminum oxide and titanium oxide was examined at the optimum conditions. All catalysts showed a promoting effect to some extent, the best result was achieved on using zinc oxide at a concentration of 2% (based on dry Cotton stalls) followed by calcium oxide and titanium oxide at a concentration of I%.

Introduction

The interest for producing chemicals from renewable resources has increased in the last decade in direct relation to the declining reserves and increasing prices of fossil fuels. Biomass residues available from agricultural processing constitute a potential source for production of chemicals such as ethanol, reducing sugars and furfural, using enzyme or acid catalyzed hydrolysis [1-8]. Furfural and its derivatives are strategic chemicals due to several possible applications. Furfural is important because it is a selective solvent for separating saturated from unsaturated compounds in petroleum refining, gas oil and diesel fuel and for the high demand of its derivatives. All pentosans containing fibrous material could in theory be used as a raw material for furfural production, however industrial production of furfural requires a minimum content of 18-20% pentosans, only about one third of the pentosans in the raw materials can be converted into furfural by means of exciting production processes.

Furfural formation from raw material containing pentosans consists of two steps, the first step involves acid--catalysed hydrolysis, where the pentoses chain are hydrolysed at high temperatures to monomeric pentose in aqueous media [9]. In a second step pentoses are converted into furfural by the elimination of water.

There are two types of technologies to produce furfural: A one stage technology where depolymerization of pentosans to xylose and dehydration to furfural occur simultaneously. And a two stage where technology isolation and depolymerization of pentosans occur under mild conditions, followed by dehydration of xylose to furfural. The advantage of two stage technology is based on the fact that the residual lignocelluloses is less degraded and can be used for conversion to other chemicals in subsequent step. The industrial process usually employed is the one stage steam production [10].

Raw materials and techniques

Raw materials: Cotton stalks were taken from local country side (El-Mansoura), hydrochloric acid, calcium oxide, zinc oxide, aluminum oxide and titanium oxide are all provided by El-Nasr laboratory chemicals, phloroglycinol was provided by Winlab laboratory chemicals. All materials were used without further purification.

Techniques: In a one litre three necked flask, a 5 gm sample (on dry basis, equivalent to 5,3 gm on wet basis) of ground cotton stalks of 0,1 mm equivalent diameter is mixed with hydrochloric acid at the specified concentration, at a certain liquor to solid ratio, then heated for the required time interval at 140[degrees]C, the temperature was automatically controlled. The produced vapours were condensed using a water condenser and the condensate was collected in a receptacle flask. Phloroglycinol was added to the condensate and left overnight in order to precipitate furfural in the form of furfural phloroglycide, which is then filtered, dried for two hours and finally weighed.

To calculate the weight of furfural, Krobber's formula [11] was used where:

Furfural = (a + 0.0052) x 0.517 a < 0.03 g Furfural = (a + 0.0052) x 0.5185 0.03 [less than or equal to] a [less than or equal to] 0.3 g Furfural = (a + 0.0052) x 0.518 a > 0.03 g

a is the weight of precipitated furfural phloroglucide

Results and discussion

A full factorial design was used to study the effect of different parameters that affect the yield of furfural [12]. In this method an upper and a lower limit of experiment conditions are chosen, for the time of hydrolysis ([Z.sub.1]) 1 hr and 4 hrs, acid concentration ([Z.sub.2]) 5 and 15% (w/w), while the liquor to solid ratio 10:1 and 50:1, different combinations between the three parameters were tested, beside three replicate experiments were performed at the centre of the design,

[Z.sup.0.sub.1] = 1 + 4 / 2 = 2.5h, [Z.sup.0.sub.2] = 5 + 15 / 2 = 10h, [Z.sup.0.sub.3] = 10 + 50 / 2 = 30h

Table (1), shows the effect of the parameters and their different combinations, as well as the replicate at the centre of the design on the yield of furfural (y).

A quick estimation shows that furfural production is favoured by rising both the time of reaction and the concentration of the acid the effect of liquor to solid ratio prevails only at low concentrations and small time of hydrolysis, at bigger values it is not significant. New dimensionless system of coordinates, (i.e: coded parameters [x.sub.1], [x.sub.2] and [x.sub.3]) are used, where:

[X.sub.J] = [([Z.sub.J] - [Z.sup.o.sub.J])/[DELTA]/.sub.J] and [DELTA][Z.sub.J] = [([Z.sub.Jmax] - [Z.sub.Jmin])/.sup.2]

The effect of these parameters and their binary and tertiary interactions on the yield of furfural are show in Table (2). A linear regression relating the yield of furfural (y) to the different parameters was derived [12], where:

y = [b.sub.o] + [b.sub.1][x.sub.1] + [b.sub.2][x.sub.2] + [b.sub.3][x.sub.3] + [b.sub.12][x.sub.1][x.sub.2] + [b.sub.23][x.sub.2][x.sub.3] + [b.sub.13][x.sub.1][x.sub.3] + [b.sub.123][x.sub.1][x.sub.2][x.sub.3]

and [b.sub.J] is given by [b.sub.J] = 1/N [N.summation over (J=1)] [X.sub.JiYi] .., so

[b.sub.o] = (+1 x 2.3 + 1 x 1 + 1 x 5.6 + 1 x 4.8 + 1 x 8.3 + 1 x 7.8 + 1 x 8.7 + 1 x 8)/8 = 5.8125 [b.sub.1] = (-1 x 2.3 - 1 x 1 - 1 x 5.6 - 1 x 4.8 + 1 x 8.3 + 1 x 7.8 + 1 x 8.7 + 1 x 8)/8 = 2.3875 [b.sub.2] = (-1 x 2.3 - 1 x 1 + 1 x 5.6 + 1 x 4.8 - 1 x 8.3 - 1 x 7.8 + 1 x 8.7 + 1 x 8)/8 = 0.9625 [b.sub.3] = (-1 x 2.3 + 1 x 1 - 1 x 5.6 + 1 x 4.8 - 1 x 8.3 + 1 x 7.8 - 1 x 8.7 + 1 x 8)/8 = - 0.4125 [b.sub.12] = (+1 x 2.3 + 1 x 1 - 1 x 5.6 - 1 x 4.8 - 1 x 8.3 - 1 x 7.8 + 1 x 8.7 + 1 x 8)/8 = - 0.08125 [b.sub.23] = (+1 x 2.3 - 1 x 1 - 1 x 5.6 + 1 x 4.8 + 1 x 8.3- 1 x 7.8 - 1 x 8.7 + 1 x 8)/8 = 0.0375 [b.sub.13] = (+1 x 2.3 - 1 x 1 + 1 x 5.6 - 1 x 4.8 - 1 x 8.3 + 1 x 7.8 - 1 x 8.7 + 1 x 8)/8 = 0.1125 [b.sub.123] = (-1 x 2.3 + 1 x 1 - 1 x 5.6 - 1 x 4.8 + 1 x 8.3 - 1 x 7.8 - 1 x 8.7 + 1 x 8)/8 = - 0.0875

Thus the regression takes the form:

y = 5.8125 + 2.3875 [x.sub.1] + 0.9625[x.sub.2] - 0.4125[x.sub.3] - 0.8125[x.sub.1][x.sub.2] + 0.0375[x.sub.2][x.sub.3] + 0.1125[x.sub.1][x.sub.3] - 0.0875 [x.sub.1][x.sub.2][x.sub.3]

The significance of the terms of the regression was tested using the student test. First the average value of the yield at the centre of the design was calculated,

[[bar.y].sup.o] = 1/N [l=N.summation over (i=1)] [y.sup.o.sub.i]

[therefore] [[bar.y].sup.o] = 1/3 (7.8 + 8 + 8.64) = 8.147

The variance at the centre of the design [S.sup.2.sub.e] was calculated where:

[S.sup.2.sub.e] = 1/N - 1 [i=N.summation over (i=1)] [([[bar.y].sup.o] - [y.sup.o.sub.i]]).sup.2],

So [S.sup.2.sub.e] = 0.143 and the standard deviation [S.sub.e] = 0.378, and the standard deviation of the jth regression coefficient [S.sub.bJ] = 0.133 where [S.sub.bJ] = [S.sub.e]/[square root of N] for the case [S.sub.bJ] Se / [square root of 8], at a significance level of 5% (95% confidence) and a degree of freedom, [[gamma].sub.2] =1. (3 - 1) 2 the tabulated value for t is t = 4.3, the calculated value is given by [([t.sub.cal]).sub.i] = |[b.sub.i]|/ [S.sub.bJ] the calculated values are as follows:

[t.sub.0] = 43.7, [t.sub.1] = 17.95, [t.sub.2] = 7.23, [t.sub.3] = 3.1, [t.sub.12] = 6.1, [t.sub.23] = 0.28, [t.sub.13] = 0.0255, and [t.sub.123] = 0.01988

The terms considered significant are those having [t.sub.cal] > [t.sub.cab]. Thus [b.sub.0] [b.sub.1], [b.sub.2], and [b.sub.12], are significant while the others are not and have to deleted from the regression which finally takes the form:

y = 5.8125 + 2.3875[x.sub.1] + 0.9625[x.sub.2] - 0.8125[x.sub.1][x.sub.2]

Now the estimated regression equation is tested to see how it fits the observations, using Fisher's test the variance ratio F = [S.sup.2.sub.res] / [Se.sup.2] where:

[S.sup.2.sub.res] = [8.summation over (1)] [([y.sub.iexp] - [y.sub.ical]).sup.2]/N - l

Where N is the number of observations and l the number of significant terms in the regression (i.e. N = 8, l = 4).

[S.sup.2.sub.res] = 1/8 -4 [[(2.3-1.65).sup.2] + [(1-65).sup.2] + [(5.6-5.2).sup.2] + [(5.6-5.2).sup.2] + [(4.8-5.2).sup.2] + [(8.3-8.05).sup.2] + [(7.8-8.5).sup.2] + [(8-9.975).sup.2] + [(8.64-9.975).sup.2]] = 1.7326 [therefore] [F.sub.cal] = 17326/0.143 = 12.116

The tabulated value of F at [[gamma].sub.1] = 4 and [[gamma].sub.2] = 2, [[upsilon].sub.1] = 4 and [[upsilon].sub.2] = 2 is 19.3, thus the regression is considered adequate because [F.sub.cal] < [F.sub.cab].

Interpretation of this regression shows that the highest value is attributed to the coefficient of [x.sub.1], followed by that of [x.sub.2], that is the increase of time of treatment and the concentration of the acid both increase the production of furfural, with the biggest share contributed to the time, the interaction term has a smaller negative coefficient, the negative sign means that at some extent the rise of both time and concentration has a deteriorating effect on the production of furfural, that may be explained by the transformation of furfural to other products, also the regression does not contain a term containing [x.sub.3], that is to say that the effect of the liquor of solid ratio is very small compared to that of time and concentration to be contained in the equation.

To find the optimum conditions for the production of furfural, the steepest ascent technique was followed by moving upon the response surface, the method suggested by Box-Wilson [12] was used, where the effect of an increment increase equal to bi [DELTA] zi/k, where k is an arbitrary constant was tested by substitution in the true regression. The true regression takes the form:

Y = 5.8125 + 2.3875 ([z.sub.1] - 2.5/1.5) + 0.9625 ([z.sub.2] - 10/5) - 0.8125 ([z.sub.1] - 2.5/1.5)([z.sub.2] - 10/5)

The first increment is [b.sub.1][DELTA][z.sub.1] = 3.58 while the second is [b.sub.2][DELTA][z.sub.2] = 4.8125, both increments are reduced by dividing each by 20, and thus they take the values of 0.175 and 0.25 respectively, we start our test from the centre of the design and we increase [Z.sub.2] by 0.175 and 0.25 respectively and see the effect of this increase on the yield, the results are listed in Table (3), 1, it is clear that the increase in extraction time and acid concentration increases the yield of the furfural, the yield exhibits a maximum value at about 3.5 hrs and a concentration of 11.5% (w/w), after which its value decreases. As a result this maximum is taken as the optimum condition, that is time = 3.5h, acid concentration = 11.5 and L/S = 30:1, the value of the liquor to solid was chosen to by 30:1, which is the reference ratio at the centre of the design as long as its term in the regression was found to be insignificant.

Three experiments were performed at the derived optimum conditions to extract furfural from cotton stalks as previously explained, the yield was 8.36%, 8.52% and 8.2% respectively with an average of 8.36% comparing this value to the value obtained from the numerical analysis the difference falls within the acceptable limits of error, it amounts to 4.1% which reveals that the estimated optimum conditions are in good agreement with the experimental results.

Effect of catalysts

Metallic catalysts have been reported as enhancers to the production of furfural, they may act as Lewis acids promoting the reaction or may stabilize the intermediates in the dehydrocyclization of pentoses. In this work four oxides namely calcium oxide, zinc oxide, aluminum oxide and titanium oxide were used. The effect of these oxides on the yield of furfural is shown in Table (4), from which, it is clear that the best improvement in the yield was achieved on using 2% ZnO where an increase of 19.6% over the blank achieved, beyond this concentration a slight decrease was noticed, that may be attributed to the formation of chloride ions on the surface of the oxide which may lower the Lewis effect. [Al.sub.2][O.sub.3] showed no effect on addition while CaO showed an increase in the yield of furfural that is independent of the concentration of the added oxide. The addition of Ti[O.sub.2] to an extent of 1% improved the yield, and an increase of 11.24% was noticed, further addition of the catalyst reduces the amount of produced furfural.

Conclusions

Production of furfural in a one-stage procedure resulted in a mild conversion of sugars, this conversion increased with increase of time and acid concentration the effect of liquor prevails only at low concentration and short time, at high values its effect was not significant, the optimum conditions for furfural production are as follows: time of hydrolysis is 3.5 hrs, acid concentration 11.5% (w/w) and a L/S = 30:1. The use of catalysts enhanced the production of furfural, the best yield was achieved on using ZnO at a concentration of 2% (based on the cotton stalks) followed by CaO and Ti[O.sub.2] at a concentration of 1%.

References

[1] Mansilla Hector D., Baeza Jaime, Urzua Sergio, Maturana Gabriel, Villasenor Jorge, Duran Nelson, 1998, Acid catalyzed hydrolysis of rice hull. Evaluation of Furfural Production, Bioresource Technology, 66, pp. 189-193.

[2] Encinar J.M., Bectron F.J., Ramiro A., Gonzalez J.F., 1997, Catalyzed pyrolysis of grape and alive bagasse. Influence of catalyst type and chemical treatment. Industrial and Engineering Chemistry Research, 36, pp. 4176-4183.

[3] Basta Altaf H., El-Saied Houssni, 2003, Furfural production and kinetics of pentosans hydrolysis in cernlobs, Cellulose Chemistry and Technology, 37, pp. 79-94.

[4] Alad S., Alonso J.L., Santos V., Parajo J.C., 1997, Furfural from wood in catalyzed acetic acid media a mathematical assessment. Bioresource Technology, 62, pp. 115-122.

[5] Al-Showiman S.S., 1999, Furfural from desert plants of Saudi Arabia Journal of the Saudi Chemical Society, 3, pp. 43-45.

[6] Al-Showiman S.S., 1998, Production of furfural from agricultural residues, Journal of the Saudi Chemical Society, 2, pp. 111-113.

[7] Al-Showiman S.S., 1998, Furfural from some decorative plants grown in Saudi Arabia, Journal of Scientific and Industrial. Research, 57, pp. 907-910.

[8] Chughtai F.A., Kamran Atif, Zill I., Huma Nazli, Nosheen S., 2002, Utilization of perocarp of peanut (Arachis hypogea) for the production of furfural and activated carbon, Pakistan Journal of Agricultural Sciences, 39, pp. 318-321.

[9] Carrasco F., Roy C., 1992, Kinetic study of dilute--acid prehydrolysis of xylan containing biomass, Wood Sci. Technol., 26, pp. 183-189.

[10] Pajtik J., Ladomersky T., Technical and economic parameter of producing furfural by a singl or two stage method, Proc. Interprogress 89 conf. Czechoslovakia., pp. 216-222.

[11] Official Methods of Analysis of American Organization of Agricultural Chemists (A.O.A.C), 8th edition, N.r. (1955), pp. 376.

[12] Akhana Zarova, S., Kafarov V., 1982, Experiment optimization in chemistry and chemical Engineering. Mir Publishers, (Moscow), pp. 151-158.

Ehab F. Abadir

Cairo University, Chemical Engineering Department

Faculty of Engineering--Cairo University

Giza--Cairo--12613, Egypt

E-mail: ehababadir@hotmail.com
Table 1: Effect of different parameters on the production of furfural.
Insert unites in table

[Z.sub.1] [Z.sub.2] [Z.sub.3] y

 1 5 10 2.3
 1 5 50 1
 1 15 10 5.6
 1 15 50 4.8
 4 5 10 8.3
 4 5 50 7.8
 4 15 10 8.7
 4 15 50 8
 2.5 10 30 7.8
 2.5 10 30 8
 2.5 10 30 8.64

Table 2: Coded values for the different parameters and their
interactions.

 [X.sub.0] [X.sub.1] [X.sub.2] [X.sub.3]

 +1 -1 -1 -1
 +1 -1 -1 +1
 +1 -1 +1 -1
 +1 +1 +1 +1
 +1 +1 -1 -1
 +1 +1 -1 +1
 +1 +1 +1 -1
 +1 +1 +1 +1
 +1 0 0 0
 +1 0 0 0
 +1 0 0 0

[X.sub.0] [X.sub.1] [X.sub.2] [X.sub.1]
 [X.sub.2] [X.sub.3] [X.sub.3]

 +1 1 1 1
 +1 1 -1 -1
 +1 -1 -1 1
 +1 -1 1 -1
 +1 -1 1 -1
 +1 -1 -1 1
 +1 1 -1 -1
 +1 1 1 1
 +1 0 0 0
 +1 0 0 0
 +1 0 0 0

[X.sub.0] [X.sub.1] y
 [X.sub.2]
 [X.sub.3]

 +1 -1 2.3
 +1 1 1
 +1 -1 5.6
 +1 -1 4.8
 +1 1 8.3
 +1 -1 7.8
 +1 -1 8.7
 +1 1 8
 +1 0 7.8
 +1 0 8
 +1 0 8.64

Table 3: The yield of furfural as calculated by the steepest
ascent method.

 Insert unites in table

Increment n% Time Conc. n Yield

 1 2.675 10.25 6.13
 2 2.85 10.5 6.449
 3 3.025 10.75 6.75
 4 3.2 11 7.049
 5 3.375 11.25 8.28
 6 3.55 11.5 8.7
 7 3.725 11.75 7.877

Table 4: Effect of catalysts on the yield of furfural.

Type of oxide ZnO [Al.sub.2] CaO Ti[O.sub.2]
Conc. of oxide [O.sub.3]

 0 8.36 8.36 8.36 8.36
 1 9 8.86 9.6 9.3
 1.5 9.6 8.87 9.5 9
 2 10 8.7 9.3 8.8
 2.5 9.48 8.8 9.4 8.63
 3 9.21 8.9 9.45 8.5
COPYRIGHT 2007 Research India Publications
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Abadir, Ehab F.
Publication:International Journal of Applied Chemistry
Geographic Code:7EGYP
Date:May 1, 2007
Words:3483
Previous Article:Adsorption of Cr(VI) and Cd(II) on charcoal from Alep pine: effects of simple chemical and thermal post treatments.
Next Article:Kinetics of dissolution of clinker components in some aggressive solutions.
Topics:


Related Articles
Low cotton prices blunt growth spike.
Effects of selected by-products of an acid hydrolyzate on cell growth and ethanol fermentation by Saccharomyces cerevisiae.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |