Synthesis of a nonpolar, chemically bonded stationary phase with low residual hydroxyl group content

Synthesis of a nonpolar, chemically bonded stationary phase

with low residual hydroxyl group content

Citation for published version (APA):

vd Venne, J. L. M., Rindt, J. P. M., Coenen, G. J. M. M., & Cramers, C. A. M. G. (1980). Synthesis of a nonpolar, chemically bonded stationary phase with low residual hydroxyl group content. Chromatographia, 13(1), 11-17. https://doi.org/10.1007/BF02302710

DOI:

10.1007/BF02302710 Document status and date: Published: 01/01/1980

Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne

Take down policy

If you believe that this document breaches copyright please contact us at:

openaccess@tue.nl

providing details and we will investigate your claim.

Synthesis of a Nonpolar, Chemically Bonded Stationary Phase with

Low Residual Hydroxyl Group Content

J. L. M. van de V e n n e * / J . P. M. R i n d t / G . J. M. M. Coenen/C. A. M. G. Cramers

Eindhoven University of Technology, Laboratory of Instrumental Analysis, PO Box 513, 5600 MB Eindhoven, The Netherlands

Key Words

Chemically bonded stationary phase Octylchlorosilanes

Residual polarity Infrared spectroscopy

Summary

A microporous silica support, LiChrosorb Si-100, has been silanized with octyldimethylchlorosilane and octyl- methyldichlorosilane. The repeatability of the silaniza- tion procedure was within about 2%. In general, these nonpolar modified silicas still contain too many residual hydroxyl groups, causing bifunctional behaviour of the adsorbent. A partial condensation of surface silanol groups at a drying temperature > 200 ~ prior to the chemical modification, decreases the residual hydroxyl group content. With respect to this residual polarity, monochlorosilanes appear to be effective. The concen- tration of bonded octyl chains remains virtually constant up to a drying temperature of 400 ~ Owing to silaniza- tion, the specific surface decreases by 1 5 - 2 0 % , whereas the pore volume decreases by 25%.

I ntroduction

The development of chemically bonded stationary phases on siliceous supports for column packing material in high- performance liquid chromatography marks an important breakthrough, n-Alkyl bonded phases especially enjoy enormous popularity. Over 60% of liquid chromatographic separations are carried out on these packing materials nowadays [1 ].

The organic molecules are bonded to the silica surface either as a polymeric layer [2, 3] or as a so-called "brush" or "bristle", a monolayer of molecules more or less perpen- dicular to the silica surface [ 4 - 6 ] . Unger et al. [7] prepared

* Author to whom correspondence should be addressed. Present address: Philips Research Laboratories, 5600 MD Eindhoven, The Netherlands.

bulk modified stationary phases by co-condensation of tetra-alkoxysilanes with organotri-ethoxysilanes. The alkyl chains become part of the bulk phase as well as of the surface.

Whatever reaction path chosen to modify the silica surface, non-modified hydroxyl groups will remain. These residual groups, including hydrolytically formed hydroxyl groups, influence the characteristics of the modified silica. The non- polar organic molecules and the polar hydroxyl groups give the surface a bifunctional character. Although these hydro- xyl groups will be deactivated by a polar eluent, their in- fluence cannot be excluded. With nonpolar eluents their in- fluence is clearly shown [8, 9].

In this paper we shall describe experiments leading to a decrease in the residual hydroxyl group concentration with maintenance of the concentration of bonded organic mole- cules. The influence of a thermal treatment of the silica before the chemical modification is examined. For the sur- face silanol group concentration for dry silica we accepted a value of 8/amole m -2. The number of bonded silane mole- cules varies from 2 - 4 / a m o l e m -~ (see Table I). Thus as long as the surface hydroxyl group concentration exceeds 4 - 5 / a m o l e m -2 after the thermal treatment, no remarkable decrease in the surface carbon content should be found, while a decrease in the residual hydroxyl group concentra- tion may be expected [10], at least when monochlorosilanes are used. In case di- or trichlorosilanes are used as modifier, all reacting bifunctionally with a yield of "~ 4/amole m -2 , a decrease of the surface hydroxyl group concentration may be disadvantageous. When the residual hydroxyl groups are located in micropores, inaccessible to the reagent, desorp- tion of surface hydroxyl groups could result in a decrease in bonded organic molecules as well.

The influence of residual hydroxyl groups is most pro- nounced with nonpolar bonded phases. As mentioned above, these phases enjoy enormous popularity. Therefore, we used nonpolar reagents, octyldimethylmonochlorosilane and octylmethyldichlorosilane, in this study. The octyl chain is a compromise between the totally bonded carbon content and a limited accessibility of the alkyl chain to micropores. After silanization, no attempts were made to decrease the influence of residual hydroxyl groups with a subsequent treatment with, for example, trimethylchloro- silane.

Table I. Surface density of alkylsilanes obtained by various authors in #mole m - 2 silylating reagent ' References ' 7 11 4.5 3.37 8 , 9 4.75 (8.7) 2.76 (4.9) 4.08 t3.5) 3.77 (2.6) 4.31 (2.9) 14 15 ' 16 3.87 - ' [ 3.34 1.90 3.49 i i 3.51 (3.84) : 3.74 ! t 3 . 5 2 ! 3.20 (3.77) 3.59 ' , 2.88 (3.04 3.47 3 14 (3-24) 3.44 ' 3.55 " trimethylchlorositane dimethyldichlorosilane allyldimethylchlorosilane t-butyldimethylchlorosilane butyldimethylchlorosilane 3.6 2.97 dibutyldichlorosilane phenyldimethylchlorosilane 2.6 octyldimethylchlorositane 3.8 2.35 octyl methyldichl orosilane 2.40 octyltrichlorosilane 2.35 decyl methyld ichlorosilane

undecylmethyldichl orosilane u ndecyltrichlorosilane dodecyldimethylchlorosilane 2.20 tridecyltrichlorosilane pentadecylmethyldichlorosilane pentadecyltrichlorosilane hexadecyldimethylchlorositane 3.4 2.36 octadecyl methyldichlorosilane octadecyltrichlorosilane heneicosyltrichlorosilane

Alkylsilane Modified Supports

The coupling of alkyl chains to the silica support can be performed with tri-, di- and monochlorosilanes. Generally, tri- and dichlorosilanes are used. The stoichiometry of these reactions has been studied extensively [11, 12].

Trifunctional chlorosilanes have an adverse influence on the residual hydroxyl group concentration. Over 50% of a di- chlorosilane should react bifunctionally if it is to be pre- ferred to a monochlorosilane. Monofunctional as well as bi- functional reactions have been found to occur when using a dichlorosilane [13]. Therefore in case mono- and dichloro- silanes are used as modifying reagents a decrease of the surface hydroxyl group concentration can be advantageous. To show the generally attainable degree of conversion, co- verages of a number ofalkylchlorosilane modified supports are presented in Table I. Collecting these values together the discrepancy in methods of calculating the surface den- sity, as recently mentioned [14], manifested itself clearly. The surface coverage should be expressed as the number of attached silane molecules instead of using the number of attached alkyl chains [8, 9]. The carbon atom of a methyl group attached to the silicon atom in mono- and dichloro- silanes should not be forgotten in the evaluation of elemen- tal analysis data [ 16 ]. The surface concentration is generally cited as #mole m -2 . Calculating this value from elemental analysis figures, one should correct for the weight increase of the support due to modification, otherwise the author will wrong himself [8,9]. Especially when long chain alkylsilanes are bonded to the silica support, approximately

30% of the total weight (for octadecylsilane 20%C) must be attributed to the bonded stationary phase.

We have used the expression given by Berendsen et al. [14] and Hemetsberger et al. [16] for calculating the alkylsilane density from carbon content data found in the literature.

%C 106

[Si] = (100 n 12 - % C M) SBE T [ # m ~ (1)

% C = weight percentage of bonded carbon n = carbon atom number of alkylsilane SBE T = specific surface of bare silica

M is the relative molecular mass of the bonded organosilane, for a monofunctional reagent given by

- S i ( C H 3 ) 2 - ( C H 2 ) n C H 3 , for di- and trifunctional silanes given by =SiX-(CH2)nCH3 or - S i O H X - ( C H z ) n C H 3 with X = C H 3 for a dichJorosilane and X = OH for a tri-

chlorosilane. The last two are supposed to react 50% mono- functionally and 50% bifunctionally.

The surface coverages, as given in Table I are calculated according to this equation. I f the values given in the articles referred to deviate, then these values are given in brackets. Table I shows that the surface coverages obtained vary be- tween 2 - 4 #mole m -2. The surface hydroxyl group con- centration was found to be 8 #mole m -2. Therefore the statement "... use a fully hydroxylated silica with the highest population of hydroxyl groups as possible ..." to achieve a dense monolayer o f bonded molecules [11 ], is a dubious one. A study of the effect of partial dehydration of the silica surface prior to silanization seems worthwhile.

Experimental

Chemicals and Materials

Three mixed 100g batches of LiChrosorb Si-100, dp = 10/am (Merck, batch EH23 charge 7518609, batch EH12 charges 7489111 and 7442231) are used as a support for the pre- paration of the chemically bonded stationary phases. Octyl- dimethylchlorosilane and octylmethyldichlorosilane are used as silanizing reagents. Because they were not commer- cially available, they were prepared by hydrosilylation o f 1-octene (Merck) and methyldichlorosilane (Fluka, Buchs, Switzerland) with potassiumchloroplatinate as catalyst [ 17-19]. After fractional distillation the reaction product was stored in closed glass ampules. Although isomers, n- octylsilane and 1-methylheptylsilane, can be formed during the reaction, only the former silane appeared to be syn- thesized as proved by NMR 2~ Toluene was used as dis- persion medium, pyridine was added to the solution to accelerate silanization. Both solvents were of Pro Analyse grade (Merck). Hexane, acetone and methanol, used to wash the modified silica were chemically pure and filtered over a 0.5/am Fluoropore filter (Millipore, Bedford, Mass. USA) before use. Water was distilled and deionized by a Millipore-Q-system. The infrared sampling procedure is described elsewhere [21,22].

The self-made chemically bonded stationary phases are compared with several commercial materials, SAS-Hypersil and ODS-Hypersil (Shandon, Runcorn, UK), Partisil ODS (Whatman, Clifton, N.J., USA), LiChrosorb RP-8 and RP-18 (Merck) and Spherisorb ODS (Phase Separations, Queens- ferry, UK). 8 ~ 6 Fig. 1 i I i I I

i

)

3

9 Weight percentage of bonded carbon as function of reaction time during silanization procedure with octyldimethyl- chlorosilane.

- Elemental analysis is applied to measure the carbon

content after modification of the siliceous support. Home made equipment was used.

- Infrared spectra of the modified silicas are recorded with a Hitachi Model EPI-G2 spectrophotometer (Hitachi, Japan). The samples are prepared according the proce- dure described earlier [21,22].

Silanization procedure

No pretreatment is given to the bare silica, except the dry- ing procedure, to avoid any change in the pore structure. Alkylchlorosilanes can be bonded to the silica surface either in the gas phase or in the liquid phase. We have appli- ed dispersion of the silica in a nonpolar solvent, analogous to the procedure o f Hemetsberger [15]. Toluene was used as dispersion liquid. A more detailed description is given elsewhere [21]. The percentage of bonded carbon as a func- tion of the reaction time is measured. The result is given in Fig. 1. Owing to the influence of the added pyridine, the reaction was complete within a few hours.

Characterization of the modified silica

Tile bare silica and the silanized silicas are characterized by several methods.

- Specific surface measurements are performed by nitro- gen adsorption at - 1 9 6 ~ according to the BET me- thod. A Str6hlein Area meter was used (Str6hlein, Dtisseldorf, FRG).

- The pore size distributions and pore volumes are calculat- ed from nitrogen adsorption/desorption isotherms using a modified Kelvin equation [23, 24]. The isotherm data were obtained with a Sorptomatic Model 1826 (Carlo Erba, Milan, Italy).

Results and Discussion

The effect o f the drying temperature on the pore structure, the carbon content and the residual hydroxyl group concen- tration is discussed.

Drying temperatures of 200 ~ 400 ~ and 600 ~ were applied. The hydroxyl group density at these temperatures has already been shown [21].

Twelve batches of octylsilane modified silica were prepared. Numbers 1 - 6 were silanized with octyldimethyl mono- chlorosilane; batches 1,2 and 3 were dried at 200 ~ 400 ~ and 600 ~ respectively, batches 4, 5 and 6 are duplicates of 1, 2 and 3. Numbers 7 - 1 2 were silanized with octylmethyldichlorosilane. Analogous to the first series, batches 10, 11 and 12 are duplicates of batches 7, 8 and 9, dried at 200 ~ 400 ~ and 600 ~ respectively.

Specific Surface and Pore Structure

As LiChrosorb Si-100 is a microporous silica, a change in specific surface is likely to occur by covering the surface with an organic layer. Results of specific surface deter- minations are given in Table II. The specific surface of LiChrosorb Si-100 is 290 + 5 m 2 g-l, in accordance with the suppliers' data. After silanization with octylchloro- silane the specific surfaces decreased by about 15-20%.

Table II. Specific surface

SBET,

LiChrosorb Si-100 LiChrosorb Si-100 dried at 600 ~ Modified silica batch 1 2 3 4 5 6 7 8 9 10 11 12 SBET Adsorbent m 2 g-1 290 + 5 280 a 290 243 240 a 230 230 a 245 2450 23O 246 240 246 255 a 242 240 a 250 260 a 252 255 246

a derived from adsorption isotherms

' Vp ' rp (dV/dr)ma x cm 3 g-1 nm nm I 1.04 7.5 7.5 i i i i 0.77 6.5 6.5 0.78 7.0 7.0 0.84 7.0 7.0 0.78 6.0 6.5 0.78 6.5 6.5 0.82 6.5 6.5

Table III. Carbon content and surface density of alkylsilanes of modified silica supports

Drying Modifier Batch temperature octyldimethyl- chlorosilane octylmethyl- dichlorosilane 1 2 3 4 5 6 7 8 9 10 1 1 12 200 ~ 40O 6O0 200 400 600 200 ~ 400 600 200 ! 400 600 %C 9.7 + 0.1 9.4 7.6 9.8 9.5 8.2 9.2+0.1 9.1 7.6 9:2 8.8 7 . 3 [Si] ~mole m -2 3.23 3.12 2.45 3.27 3.16 2.67 3.41 3.37 2.74 3 . 4 1 3.24 2 . 6 2

The differences between the various batches are not signi- ficant. The BET surfaces derived from the adsorption iso- therms are somewhat lower. The specific surface decrease is partially due to the weight increase by silanization, partially due to the filling of micropores by the organic molecules. The decrease cannot be attributed to a surface reduction owing to a temperature effect, since the specific surface of the bare silica remained unchanged after drying at 600 ~ More insight is given by data about the pore structure. Data are given in Table II. The repeatability of the equip- ment, used to measure the nitrogen adsorption isotherms, is 3 - 4 % . The variations between the data of the modified

silicas are slightly higher but they are not.so significant that conclusions from differences between the three drying temperatures or the silanizing agents can be drawn. The LiChrosorb Si-100, which we used, had an average pore radius of 7.5 nm. Table II shows that after silanization the pore volume has been decreased about 25%. Consequently, the average pore radius decreased from 7.5 nm to 6.5 run.

Surface Coverage

In Table III the weight percentages of bonded carbon and the subsequent alkylsilane densities are listed. Conversion

I00 50 0 I00 .s u~ u~ 50 0 I00 50

/

Infrared spectra of octyIdimethylmonochlorosilane modifi- ed supports, dried at different temperatures prior to modi- fications, visualizing residual hydroxyl group concentration.

100 50 0 100

5

i~ 5o

9 Infrared spectra of octylmethyldichlorosilane modified supports, dried at different temperatures prior to modi- fication, visualizing residual hydroxyl group concentration.

rates are not given as the stoichiometry of reactions with bifunctional silanes is not yet clearly established. The surface concentration of bonded silane molecules is in accordance with the values given in Table I. Eq. (1) is applied. The elemental analysis data are accurate within 2%. The re- peatability of the silanizing procedure appears to be good. Drying of the siliceous support at 400 ~ has not led to a significant decrease in the carbon coverage. There were enough attainable silanol groups left at the surface to achieve the maximally attainable conversion. As predicted before [21,22], drying at 600 ~ induced a decrease in the carbon coverage. A reduction of 1 5 - 2 0 % is found. But even this decrease is rather small since from the literature [25]. It can be estimated that the silanol group density decreases from 8/amole m -2 to 2 . 3 # m o l e m -2 after

drying at 600 ~ The differences between octyldimethyl- chlorosilane and octylmethyldichlorosilane are hardly significant.

Residual Hydroxyl Group Concentration

The residual hydroxyl group concentration has been visualized by infrared spectroscopy, using the mull tech- nique as described elsewhere [21, 22], From batches 1, 2 and 3 and 7, 8 and 9 mull samples were prepared and infra- red spectra were recorded. The samples were dried at 200~ during 4 hours prior to embedding them in the mull oil. Thereby all physisorbed water was evaporated. Results are given in Figs. 2 and 3. The spectra of both octyldimethyl- chlorosilane modified silicas and octylmethyldichlorosilane

modified silicas reveal that the residual hydroxyl group concentration markedly decreases with increasing drying temperature.

A slight difference can be noticed between Fig. 2 and 3. The shoulders in the OH-absorption band around 3550cm -1 are somewhat broader for the dichlorosilane modified packings, owing to formation o f new hydroxyl groups by hydrolysis of nonreacted chloride atoms. These hydroxyl groups can be hydrogen bonded to silanol groups on the silica surface.

For comparison, infrared spectra were recorded from a number of commercially available alkylsilane modified packing materials. The spectra are represented in Fig. 4. The modified silicas are dried at 200 ~ prior to embedding

them in the mull oil. Unfortunately it is not known whether these packing materials have been treated with a sman reactive silylating agent, like trimethylchlorosilane, to decrease the residual hydroxyl group concentration. The spectra visualize the number of residual hydroxyl groups. Most modified silicas exhibit a more or less equal residual hydroxyl group concentration. Differences in the absorp- tion frequency around 3550 cm -1 are most pronounced. Partisil ODS contains more residual hydroxyl groups, which can be explained by its low carbon content of 5%. The Hypersil materials contain fewer residual hydroxyl groups. Our samples dried at 400 ~ and 600 ~ exhibit a smaller hydroxyl group concentration as compared to most of the packing materials shown in Fig. 4.

tO0 50 0 100 01 o l ~ so 0 tOO 50 L i c h r o s o r b R P - 8 i I I I I I I , I Lichrosorb RP-18 I I I I I i i I I I I Fig. 4

\

o;/

\

Infrared spectra of number of commercially available alkylsilane chemically bonded stationary phases, dried at 200 ~ prior to embedding them in mull oil.

Conclusions

Drying o f the silica sample at such a t e m p e r a t u r e that not only all physisorbed water is desorbed but also part o f the surface silanol groups are condensed, has a favourable effect on the residual h y d r o x y l group concentration. I f the concentration o f b o n d e d alkyl chains is to be constant, a temperature o f 400 ~ seems appropriate. In case a small reduction in the carbon c o n t e n t is permissible, the drying temperature can be further increased. Monochlorosilanes are better than dichlorosilanes.

Owing to the modification, the specific surface and the pore size decrease b y 15% and 25%, respectively. The dry- ing temperature has no influence on the specific surface and the pore structure o f the m o d i f i e d packing materials. Chromatographic measurements in reversed-phase systems and normal phase systems will give additional information on the influence o f residual p o l a r i t y o f nonpolar m o d i f i e d supports.

Acknowledgement

A grant from DSM Research, Geleen, The Netherlands, to one o f us (J. v . d . V . ) is gratefully acknowledged.

References

I l l R.E. Majors, J. Chromatogr. Sci. 15, 334 (1977).

12] J . J . Kirkland, J. J. DeStefano, J. Chromatogr. Sci. 8, 309 (1970).

131 J..s Kirkland, J. Chromatogr. Sci. 9,206 (1971).

141 R. E. Majors, M. J. Hopper~ J. Chromatogr. Sci. 12, 767 (1974).

[51 D.C. Locke, J. Chromatogr. Sci. 11,120 (1973). 161 V. Rehhk, E. Srnolkova, Chromatographia 9,219 (1976).

[7] K. K. Unger, N. Becker, P. Roumeliotis, J. Chromatogr. 125, 115 (1976).

18l K. Karch, L Sebastian, L Halasz, J. Chromatogr. 122, 3 (1976).

[91 K. Karch, Thesis, University of Saarbriicken, 1974.

[10l R. P. IV. Scott, P. Kucera, J. Chromatogr. Sci. 13, 337 (1975).

[11] P. Roumeliotis. K . K . Unger, J. Chromatogr. 149, 221 (1978).

[121 K. Berg, K. Unger, Koll. Z. u. Z. Polymere 246,682 (1971). [131 M . L . Hair, J. Colloid Interface Sci. 60, 154 (1977).

1141 G. E. Berendsen, L. de Galan, J. Liquid Chromatogr. 1,561 (1978).

1151 H. Hemetsberger, W. Maasveld, H. Ricken, Chromatographia 9, 303 (1976).

[161 H. ltemetsberger, M. Kellermann, H. Ricken, Chromato- graphia 10, 726 (1977).

[171 E. W. Pietrusza, L. H. Sommer, E. C. Whitmore, J. Am. Chem. Soc. 70, 484 (1947).

1181 P. Tarrant, G. W. Dyckes, R. Dunmire, G. B. Butler, J. Am. Chem. Soc. 79, 6536 (1957).

[ 19] E. Y. Lukevits, M. G. Voronkov, Organic Insertion Reactions of Group IV Elements, Consultance Bureau, New York, 1966. [20] G. J. M. M. Coenen, Graduation Report, Eindhoven Uni-

versity of Technology, 1979.

[21] J. L. M. van de Venne, Thesis, University of Technology Eindhoven, 1979.

[221 J. L. M. van de Venne, J. P. M. Rindt, G. J. M. M. Coenen,

J. Colloid Interface Sci., in press.

I231 S. J. Gregg, K. S. W. King, Adsorption, Surface Area and Porosity, Academic Press, New York, 1967.

1241 Z C. P. Broekhoff, B. G. Linsen, in B. G. Linsen (ed.), Physical and Chemical Aspects of Adsorbents and Catalysts, Academic Press, New York, 1970, Chapter 1.

[251 V. Ya. Davydov, A. V. Kiselev, L. T. Zhuravlev, Trans. Faraday Soc. 60, 2254 (1964).

Received: Aug. 23, 1979 Accepted: Aug. 24, 1979 D