Adsorption of Methylene Blue on Activated Carbon An Experiment Illustrating Both the Langmuir and Freundlich Isotherms J. H. Pofgieler Division of Physical Metallurgy, Council for Mineral Technology, Private Bag X3015. Randburg. Republic of South Africa Adsorption and adsorption processes are important fields of study in physical chemistry. They form the basis for understanding phenomena such as heterogeneous catalysis, chromatomanhic analvsis.. dveine .. . .. of textiles. and clarification of various effluents.'~cti\.atedcarbon finds particular application in the rlarificntion of effluents, includine the removal of coloring matter from various types of solution and the elimination of organic suhstances from treated Dotable water (especially Gihalomethanes) and waste water. However, findine simple and easilv ~erformahleex~erimentsto illustrate the quantitative as&& of adsorption can be very difficult. A number of such experiments were described in some previous issues of this Journal (1-5). The present investigation describes the adsorption of methylene blue, a rationic organic dyestuff commonly used for trarer studies in water rrsearch, on activated rarbon. Althoueh it has been claimed (2) in a orevious investienti(m using Gfferent adsorhents, t h a t meth;lene blue a d s i p t i o n ohevs the Frumkin tvDe of adsor~tionisotherm. it was found in the present case tiobey hoth'the ~ a n ~ m u i r ' a nFreundd lich isotherms better for the specific type of adsorhent used.
.
Experlmental Procedure The adsorbent used in the experiment was a commercial activated carbon, the physical characteristics of which are given in Table 1.A methylene blue solution with a concentration of 25 mgL was prepared from analytical-grade reagent and distilled water. The posiof this solution was determined tion of maximum absorbance (A,,) to be at 630 nm on a Speetronic 20 spectrophotometer. Coneentrations of 525 mgL of this methylene blue solution were found to obey the Beer-Lambert law. Adsorption isotherms were obtained as follows: Accurately weighed samples of between 0.001 g and 0.100 g of adsorbent were olaced in seven seoarate 250-mL conical flasks. each containine 100
logr =logK+%,logc
(2)
where K, n = constants and x, c have the same meaning as in theLangmuir isotherm. I t is clear from eq 2 that aplot of log x against log c should yield a straight line with a slope of 'I, and an intercept of log K. The equation describing the Frumkin isotherm can be written as u=
log (0155.55) + 2aOl2.303
(3)
where" = log (8(1- 8).c), 0 = M/M,d,, M = the amount of dye adsorbed at equilibrium, Mads= the maximum amount of dve adsorbed a t equilibrium (2.48 mg in this experiment), and c ~ , d= constant;, whilec has rhesame meaning as before. Plotting the first member against 0 should yield a straight line -if the Frumkin isotherm is oheved. From the interce~t ---~. with the ordinate 0 is obtained wgile a can be calculated from the slooe. The experimental data points ~ummarizedin'l'ahle 2 were fitted toall three isotherms, which a r e r r a ~ h i r a lrepresenrl~ ed in Figures 1-3. The dotted line in e&hfigure iepresents a linear reeression fit to the results. However, since adsorption data are of a nonlinear nature, nonlinear regression was also performed on each set of data points. These nonlinear regression fits are represented as solid lines in Figures 1-3. The correlation coefficients obtained with both kinds of regression for all three adsorption isotherms are summarized in Table 3. Except in the case of the Langmuir isotherm, nonlinear regression fits produced better correlation coefficients and thus more accurate constants than are obtained with the ~
~~
~
~
Table 1. Physical Properlles of the Adsorbent ical shaker for -72 h to reach equilibrium. At the end of this time analysis of the supernatant solution in each flask was carried out by spectrophotometry at 630 nm. All the experimental runs were conducted at 25 "C.
BET surface area, m2/g Pore vol~me.em3 lodlne number (mglg) Apparent denshy, g/cm3
Effective partide size, mrn Abrasion number
Treatment of Results The derivation of the Langmuir, Freudlich, and Frumkin adsorption isotherms is dealt with in most physical chemistry textbooks. It will therefore suffice to state only the final form of each equation in this discussion. The Langmuir adsorption isotherm may he written as
where x = amount of solute (MB) adsorbed per mass of adsorbent (act.C), x, = limiting amount of adsorbate that can be taken up per mass of adsorbent,K = a constant, and c = concentration of the solute in the solution that is in equilibrium with the adsorbent. Thus the plot of 1/x against l l c should be linear with agradient of l/x,Xand intercept of I / x, on the 1/x asis. The equation describing the Freundlich isotherm can he defined as
Table 2.
950-1050 0.8-0.9 900-1000 0.460 0.8-1.1 75-80
Experlrnental Adsorptlon Data
Activated carbon
lniiiai h4E mass before
Final blB mass after
mass (mg)
adsorption
adsorption
(me)
(mg)
Volume 68 Number 4
[w Mass of MB adsorbed (mg)
April 1991
after
adsorption (mg/dm3)
349
-
F W ~1. Application of me F ~ m k i n e q u to ~ ~me o ~experlmentai datawinls determined lor lha adsorption 01 melhylene blue on an activated carbon at 25 OC.
40.0
Table 3. Correlallon Coelflclentsfor Llnear and Nonlinear Regresalon Fllr ol Dlflerent Adsorption Isolhermr
35.0 3 0 . 0
-
25.0 20.0
Figure 3. Application 01 lha Freundlich equation to the experimental data points determined iwme adsorption of rnethylene blue on an activated cmon at 25 OC,
-
-
aasorptlon isolherm
Carelation Linear regression
Coefficient Nonlinear regression
Frumkin Langmuir Freundlich
0.940 0.998 0.987
0.984 0.998 0.992
Type of
15.0 -
O0.0 0.0
1.0
2 . 0 , 3 . 0
4.0
5.0
6.0
Figure 2. Application of the Langmuir equation to Ite experimental data poinls determined lor the adswptlon of methylene blue on an activatsd carbon at 25 'C.
Table 4.
Langmulr and Freundllch Isotherm Constants
T y p e d regression
Linear Nonlinear
k
Langmuir
0.378 0.355
K
0.358 0.401
Fremdllch K n 0.0834 0.0897
1.66 1.50
ordinary linear regression. From these values i t is clear that although the adsorption of methylene blue on Pyrex beads and Taz06 can be described by the Fmmkin isotherm (2), this is not the case for activated carbon. The correlation coefficient values in Table 3 also indicate that the data fit the Langmuir isotherm better than the Freundlich isotherm, both in the case of linear and nonlinear regression. I t is furthermore evident that both the Langmuir and Freundlich isotherms describe the adsorption process very well and are better adhered to than the Frumkin isotherm. The values of the constants obtained for the Langmuir and Freundlich isotherms from both types of regression are given in Table 4.
where N = Avogadro's number and a = the area of the dye molecule. In this case the value of "a" for methylene blue can be taken as 120 AZ(6). This simple experiment brings to the student's attention the basic principles of adsorption as well as the application of more than one isotherm to describe the whole process. It also enables the student to judge the suitability of different types of regression fits to experimental data, to become acquainted with the measurement of specific surface area of the adsorbent, and to estimate an approximate value for it.
Calculation ol the Specllk Surlace Area of the Activated
Uterature Cited
Carbon
I. ~ ~ fD f G, . : R ~ ~ ~ , S . M . C . ; V ~ U ~J .~c ~h ~e r, nD. ~. H d u. c1%8,65,815-816. .
350
Journal of Chemical Education
S = x,.N.a
(4)
