Characterization of Specific Interactions Capacity of Solid Surfaces by Adsorption of Alkanes and Alkenes. Part II: Adsorption on Crystalline Silica Layer Surfaces G. Lignerl*/M. Sidqit/j. Jagiellol'2/H. Balardl/E. Papirer 1 Centre de Recherche sur la Physico-Chimie des Surfaces Solides, C.N.R.S., 24, avenue du Pdt Kennedy, F-68200 Mulhouse, France 2Visiting scientist. Institute of Energochemistry of Coal and Physicochemistry of Sorbents, University of Mining and Metallurgy, PL-30-059 Krakdw, Poland

ways sufficient to describe the specific interaction capacity of the surface. Additional informations can then be obtained by AH~ and ASh calculations. The purpose of the present paper is to show how this method works when the probes are adsorbed on surfaces of crystalline layer of phyUitic silicas.

Key Words Gas solid chromatography Crystalline and amorphous silicas Alkanes and alkenes :Effects of crystal structure

Materials and Methods Summary

Materials

Adsorption of branched octanes and linear hydrocarbons on crystalline lamellar silica surfaces has been studied by inverse gas chromatography at infinite dilution. Taking the adsorption of the n-alkanes as a reference, the influence of the double bond on the hydrocarbon adsorption phenomena has been demonstrated. Thermodynamical parameters have been calculated which permit conclusions to be made on the adsorption mechanisms of lamellar materials.

Three silicas were used: - A fumed silica, Aerosil 200 (Degussa, 200 m2/g.): sample A - A precipitated silica, Zeosil 175 MP (Rh6ne Poulenc 180 m2/g.): sample P - A crystalline lamellar silica, H-kenyaite (20 m2/g.): sample L

Probes

Introduction Many studies in the literature have been devoted to the surface properties of silicas [1-3] since these properties are important for applications such as a reinforcing filler, catalyst, chromatographic support etc. Among the techniques leading to the determination of the surface thermodynamic parameters and especially the surface energy, adsorption methods are widely used [4-6]. However, studies have been carried out mainly on amorphous macroporous silicas, with so called "open surfaces". In the first part of this work [7], we presented the inverse gas chromatography method, at infinite dilution, which permits an assessment of the non-specific and specific surface interaction capacity of the solid surface using linear alkane and alkene as probes. It was seen that the comparison between the free enthalpies of adsorption of the saturated and unsaturated hydrocarbons allows an increment of energy e~, linked to the presence of the double bond to be defined. However, the knowledge of ~, alone is not alChromatographia Vol. 29, No. 1/2, January 1990 0009-5893/90/1 0035-04 $ 03.00/0

The probes used (methane, n-alkanes, n-alkenes,. branched octane isomers ...) were purchased from Aldrich as puriss grade products and were used without any further purification.

Inverse Gas Chromatography (IGC) Silica particles, in the 250 to 400 ~tm range, for the filling of chromatographic columns were obtained as described in Part I [7]. Silica L particles could be sieved directly from the initial product. The column was attached to a gas chromatograph (IGC 120 DFL from DELSI), fitted with a flame ionization detector. Helium was used as carrier gas with a flow rate of approximately 20 cm3/mn. Injector and detector temperatures were set at 100 ~ i.e., a temperature above the measuring temperature. The silica samples were conditioned at 160 ~ in the chromatographic column under helium gas flow. Very small volumes of gaseous solutes were injected so as to approach conditions of gas chromatographic linearity. Usually, symmetrical peaks were recorded.

Originals

9 1990 Friedr. Vieweg & Sohn Verlagsgesellschaft mbH

35

For non symmetrical peaks, the retention times were determined from the first order moment method using a Shimadzu C R 3 A integrator and our own specific computer program. Corrected retention volumes were taken for the calculation of thermodynamic quantities.

Table I Relative values of enthalpies of adsorption branched isomers of octane against n-octane (AHcs/A = 66.5 kJ/Mol; AHcs/P = 65.9 kJ/Mol; AHc8IL 84.1 kJ/Mol). Probes

(AHi - AHcs)/AHcs A

Results and Discussion The surface properties of silicas have been intensively studied for many years and more recently, inverse gas c h r o m a t o g r a p h y at infinite dilution t e c h n i q u e has b e e n often e m p l o y e d [8]. The main characteristics which are o b t a i n e d from these c h r o m a t o g r a p h i c measurements are the dispersive c o m p o n e n t of the

Octane 2-Methylheptane 2.5-Dimethylhexane 2.2-Dimethylhexane 2.2.4-Trimethylpentane 2,3.4-Trimethylpentane Hexamethylethane

0.00 % 0.25 % 2.94 % 2.75 % 5.56 % 3.36 % 1.39 %

P

L

0.00 % 4.08 % 5.91% 6.48 % 12.55 % 6,91% 11.46 %

0.00 % 16.79 % 35.65 % 35.55 % 37.11% 35.65 % 41.93 %

surface energy 7 D and the contribution due to specific interactions 1 ~p. 7 sD can be easily calculated using the method proposed by Dorris and Gray [9], but the specific c o m p o n e n t of the surface energy can only be estimated by an I~p parameter as shown by Saint-Flour and Papirer [10] and by Schultz et al. [11]. In the first part of this work, a new method for the determination of specific interaction capacity was proposed by extending the Dorris and Gray theory using alkene probes [7]. When reporting the free enthalpy of adsorption of the h y d r o c a r b o n probes versus their number of carbon atoms, a 1-alkene line was obtained parallel to, but above the n-alkane line. Thus, the characteristic energy increment, due to the presence of the double bond, was called en. The meaning of e~ as a characteristic parameter of the double bond, was reinforced by the results obtained by the injection of dienes. However, this study was devoted to amorphous fumed and precipitated silicas only. In a previous paper, a comparative study of the dispersive components of the surface energy of silicas of various origins was presented [12]. A characteristic behavior was evidenced either for amorphous or for crystalline silicas. It was shown that the adsorption phenomena are different on the two types of material and that this difference of adsorption can be demonstrated using alkane isomers. In Table I we r e p o r t the calculated values of the enthalpy of adsorption of n-octane and its branched isomers. From the values of AH for the branched compounds relative to n-octane, it is seen that all isomers interact quite similarly with the surface of the A and P amorphous silicas. However, the difference between the values of A and P suggests that the adsorption on p r e c i p i t a t e d silica d e p e n d s m o r e on the steric hindrance of the probe than it does in the case of the fumed sample. This results is in a good agreement with the observation made in part I about the difference of the silanol accessibility between the two silicas. In the case of the crystalline silica L, the relative values of AH show that the interactions of the surface with the branched isomers are much lower than with linear octane. This results can be explained in the knowledge that the crystalline structure of H-kenyaite is composed by superposed layers which leave a basal spacing distance of 15.6 A. Subtracting the silanol

36

group size on both sides of the interlayer space, or about 5 A remain. Thus, the n-alkane probe can r netrate between the layers of the crystalline structu: interacting with both sheets, while the branched is mers are excluded from the basal space and ads~ only on the external surface. It was then interesting to verify if our method of termination of the specific interaction capacity usi: alkene probes works with silica L. Saturated and saturated linear hydrocarbons were injected in the lumn containing sample L. In Figure 1, we show lq Ln Vn versus the number of carbon atoms no, of t: probes. As observed in part I, the n-alkane plots sh( the traditional straight line while a curve is obtain: with the 1-alkene probes, coming closer to the alka: line for higher values of no. This means that the ener increment e~ is changing with n~. In addition, enthal of adsorption of the probes on silica L were calculat: and in Figure 2, we show the variation of the AH w sus n~. The shape of the curves are similar in the ca of RT Ln Vn or AG ~ These results suggest that the sorption of 1-alkenes resembles that of alkanes wh~ the length of the molecule increases and the relati influence of the double bond decreases. This is co trary to observation made on open surface silicas, Part I.

25

ka/moi

RT Ln (Vn) at IOOOC 20

3.../j..i..

15'

o~ j.~ I ~ 0

[ * Alkanes [ o Alkenes I x Dienes

"

I

I

I

I

5

6

7

8

Figure 1 Variation of net retention volume of linear hydrocarbons crystalline silica L. versus their number of carbon atoms.

Chromatographia Vol. 29, No. 1/2, January 1990

Origim

To explain this surprising behavior, we may consider the mechanism of adsorption of linear hydrocarbons between the layers of crystalline silicas. From Table II, it is seen that the enthalpy of adsorption AH is higher in the case of the crystalline silica L than for -the amorphous silicas A and P. -This indicates that the molecules are strongly held _between the layers of silica L and have reduced mobility along the surface. It follows that the longer the chain is, the more difficult it is to find the interaction between the OH groups and the ~ bond, a situation which could explain the diminishing of e~. In Figure 1, we also show RT In Vn or AG~ of adsorption of linear or-co dienes on silica L. It is seen that for -these probes, the plot is below the alkane line which means that the adsorption of hydrocarbons becomes more difficult when there is a double bond in the molecule. The explanation of these results can be given on the basis of conformational possibilities of :the probes. From the model discussion of De Gennes [13], it is seen that the rotation of molecule fragments around the a bonds could be the driving force of the insertion between the sheets. Comparing both ends of the 1-alkene molecule. Figure 3 shows that the rotation around the first o bond produces a larger dynamic steric hindrance for the unsaturated than for the saturated end. It follows that the adsorption of alkenes should start from the saturated end. Obviously, this ihypothesis explains the lower AG~ and AH of adsorp:tion of dienes. Thus, the adsorption of these latter :molecules takes place on the external surface and :therefore the AH is similar to that observed with amorphous silicas A and P (Table II).

II

Unsaturated end of a l k e n o

Figure 3 Dynamic steric hindrance of alkcne molecule due to the rotation around the first o bond.

Table II Enthalpies of adsorption of linear hydrocar-

bons on silica A, P and L Probes

Conclusion Using saturated and unsaturated hydrocarbons as inverse gas chromatography probes, phenomena peculiar to adsorption on lamellar materials are evidenced. Comparing the adsorption of branched and linear alkane isomers, it is seen that n-alkane molecules penetrate the space between the superposed crystalline

90 80

'AI4ofadsorption (kJ/mol)

o~ /

70

/~

60

ik.-g;ff ;

j

50

o

Saturotod ond o f o l k o n e

AH (kJ/mol) A

P

L

Pentan e Hexane Heptane Octane

20.6 25.6 33.0 66.5

33.2 41.8 55.0 65.9

40.7 54.8 70.5 84.1

Pentcnc Hexene Hcptenc Octcne Hexadiene Heptadiene Octadiene

39.1 44.1 51.4 57.0 62.6 69.9 75.5

40.1 48.7 61.9 72.8 55.6 68:8 79.7

48.3 60.6 73.1 85.0 52.1 63.1 71.3

layers of silica L. Moreover, a decrease of the specific interaction capacity of the surface, described by the En parameter, with increasing number of carbons of the probes, can be observed. This phenomenon is explained by steric hindrance effects, due to the higher rigidity of the unsaturated molecules involved in the adsorption mechanisms. The present study shows clearly that in some particular cases the method proposed in Part I of this work cannot be applied in a straightforward manner. In all instances, however, interesting information can be extracted from inverse gas chromatography experiments especially concerning the surface topology of solids.

Alkenes

"41:1

References n

.30

i

i

i

i

5

6

7

8

Figure 2 Variation of enthalpy of adsorption (AH) of alkanes and alkenes

on crystallinesilica L. versus their number of carbon atoms. Chromatographia Vol. 29, No. 1/2, January 1990

[1] II. Balard, M. Sidqi, E. Papirer, J. B. Donnet, A. Tuel, 11. tIornrnel, A. P. Legrand, Chromatographia, 25, (8) 707 (1988). [2] A. Vidal, M. J. Wang, E. Papirer, J. B. Donner, Chromatographia, 23, 121 (1987). [3] M. ZaborskL A. Vidal, G. Ligner, IL Balard, E. Papirer, A. Burneau, Langmuir, 5, (2) 447 (1989).

Originals

37

[4] D. M. Young, A. D. Crowell, Physical Adsorption of Gases, Butterworths, London, 1962. [5] A. V. Kiselev, Ya. L Yashin, Gas Adsorption Chromatography, Plenum Press, New York, 1969. [6] R. M. Barrer, Zeolites and Clay Minerals as Sorbents and Molecular Sieves, Academic Press, London, 1978. [7] M. Sidqi, G. Ligner, J. Jagiello, 1L Balard, E. Papirer, Chromatographia, 28,588 (1989). [8] E. Papirer, A. Vidal, 11. Balard, in "Inverse Gas chromatography", ACS Symposium Series 391, chap. 18 (1989). [9] G. M. Dorris, D. G. Gray, J. Colloid Interface Sci., 71, 93 (1979).

38

[10] C. Saint-Flour, E. Papirer, J. Colloid Interface Sci., 91, (1) i (1983). [11] J. Schultz, L. Lavielle, C. Martin, J. Chimie Physique, 84, (i 231 (1987). [12] G. Ligner, A. Vidal, H. Balard, E. Papirer, J. Colloid lnte face Sci., 133, (1) 200 (1989). [13] P. G. De Gennes, Scaling Concept in Polymer Physic Cornell University Press London (1979). Received: Sept. 28, 198~ Accepted: Oct. 18, 1989 E

Chromatographia Vol. 29, No. 1/2, January 1990

Original