matography,” Marcel Dekker, New York, 1965. (9) Giddings, J. C., J. Chromatog. 5, 61
(19611. (10) Ibid., 13, 301 (1964). (11) Giddings, J. C., Robison, R. A., ANAL.CHEM.34, 885 (1962). (12) Golay, M. J. E., “Gas Chromatography 1958,” D. H. Desty, ed., p. 36, Butterworths, London, 1958. (13) Gordon. S. M.. Kriee. G. J.. Haarhoff, ’P. C., Prktoriu; V., ANAL: CHEY.35, 1537 (1963). ~
(14) Jones, W. L., Zbid., 33,829 (1961). (15) Knox, J. H., McLaren, L., Zbid., 35, 449 (1963). (16) Ibid., 36, 1477 (1964). (17) McLaren, L., Ph.D. thesis, Edinburgh, 1965. (18) Perrett, R. H., Purnell, J. H., ANAL. CHEM.35, 430 (1963). (19) Sternberg, J. C., Poulson, R. E., Zbid., 36, 1492 (1964). (20) Taylor, G. I., Proc. Roy. SOC.A219, 186 (1953). (21) Ibid., A223, 446 (1954).
(22) Terry, W. M., Blackwell, , R. J., Rayne, J. R., quoted by J. C. Giddings, ANAL.CHEM. 35, 1338 (1963). RECEIVEDfor review June 25, 1965. Accepted September 7, 1965. Third International Symposium on Advances in Gas Chromatography, University of Houston, Houston, Tex., October 18-21, 1965. The author expresses his gratitude to the National Science Foundation for the award of a Senior Visiting Scientist Fellowship at the University of Utah,
where the work was carried out.
Electrostatic Interactions in Gas-Solid Chromatography JAMES KING, Jr. Chemistry Section, Space Science Division, Jet Propulsion laboratory, California Institute o f Technology, Pasadena, Calif.
SIDNEY W. BENSON Department o f Kinetics and Thermochemistry, Stanford Research Institute, Menlo Park, Calif. An electrostatic theory of physical adsorption i s applied to gas-solid chromatography and predicts the elution order of many gases from columns. When applied to the separation of the rare gases and methane on a y-AI2o3 column at room temperature the theory not only gives the correct elution order, but it also quantitatively explains the behavior of methane on the column. The theory gives a qualitative explanation of the chromatographic inseparability of argon and oxygen at room temperature. The application of the theory to molecular sieve column leads to the suggestion that adsorption on molecular sieves i s very similar to adsorption on other adsorbents like y-AI2o3and silica gel. The theory proposes that adsorption on all of these adsorbents i s governed by electrostatic interactions. This concept when applied to sieve columns leads to refutation of the accepted hypothesis that the separation of molecules on the column i s caused b y some type of “sieve” action.
M
of adsorption in Gas-Solid Chromatography (GSC) have attributed the gas-solid interactions to non-polar van der Waals’ forces. This concept, which was used by Langmuir ( I S ) to explain monolayer adsorption and by Brunauer, Emmett, and Teller (5) to explain multilayer adsorption, is based on the assumption that the same forces that produce condensation are responsible for the binding energy of molecules to surfaces. The arguments which are presented to support this hypothesis are : (a) Physically absorbed layers, particularly those many molecular diameters thick, behave in many resects like two-dimensional liquids. (b) Physical adsorption, like condensation, is a general phenomenon and will occur with any gas-solid system provided only that the conditions of temperaOST OF THE TREATISES
ture and pressure are suitable. (c) Since physical adsorption is related to the process of liquefaction, it occurs to an appreciable extent a t pressures and temperatures close to those required for liquefaction and, under suitable conditions, physically adsorbed layers several molecular diameters thick are formed. (d) The heat of physical adsorption is of the same order of magnitude as the heat of liquefaction. None of these arguments, except possibly (d), relate directly to the interaction of the gases with the surface. The interaction between adsorbed molecules can involve non-polar van der Waals’ forces without these forces being primarily responsible for the interaction between individual molecules and the surface. The adsorption of gases on a chromatographic column generally occurs at temperatures much higher than the temperature required for liquefaction. For example, the hydrogen isotopes (9) are separated on an A1203 column a t 77.4’ K. although their boiling points are between 20’ K. and 30” K. Also, the separation of the rare gases (Figure 1) on an alumina column occurs a t room temperature although the range of their boiling points is from
Table I.
27” K. for neon, to 166’ K. for xenon.
If the heats of adsorption and the heats of liquefaction were essentially equal, this would indeed imply that the same forces were responsible for both processes. However, for many gases, especially the permanent and inorganic gases, the heats of adsorption are two to three times larger than the heats of vaporization. This is seen in Table I where the heats of adsorption of a number of gases on various adsorbents are listed along with the corresponding heats of vaporization. The largest difference between AH,,, and is shown by the hydrogen isotopes. The magnitude of the heats of adsorption of the isotopes on alumina was one of the factors which led King and Benson (10) to postulate that electrostatic forces were responsible for the adsorption of the isotopes on alumina and other adsorbents. THEORY
According to the electrostatic model of adsorption developed by King and Benson, gases are polarized by strong surface electric fields and by the uncompensated charges on the surface
Heats of Vaporization and Heats of Adsorption of Gases on Various Adsorbents
Adsorbent ( - AHada kcal./mole) AHvap
Gas Ar 0 2
Nz Kr
co
CH4 CzHe
coz
CsHs p-Hz o-Hz HD 0-Dz p-Dz
(kcal./mole) 1.50 1.63 1.33 2.16 1.41 2.21 2.25 3.84 3.67 0.22 0.22 0.26 0.29 0.29
Carbon
Silica gel
Alumina
Molecular Sieve 5A
2.7 2.7 2.9 3.4 4.8
1.5 2.0 4.2 3.2 2.8 5.5 6.3 8.0
VOL. 38,
2.5 4.7
4.0
7.2 1.40 (77.4’ K.) 1.55 (77.4OK.) 1.51 (77.4’ K.) 1.58 (77.4OK.) 1.65 ( 7 7 . 4 ’ K . )
NO.
2, FEBRUARY 1966
0
261
4.0
surface fields of the order of 100,000,000 volts per cm. The fields were strongest over an Al+3 site and a vacancy site in the A1203 structure. They attributed these strong fields to the existence of normal structural vacancies in the lattice of the adsorbent. All defect structures such as 3-2 or 3-4 lattices must have these vacancies in order to maintain charge neutrality. The field over the Al+3 site was positive and that over a vacancy was negative. The positive and negative sites could explain the origin of the acidic and basic surface sites which have been postulated for alumina and silica-alumina catalysts
2.0
(4)'
1.0
.-Ec
*
0.6
0.4
0.2
0.I a, A.8
Figure 1 . Retention times of the rare gases and methane on an AI2O3column as a function of their respective polarizabilities
as they approach the surface, This electrostatic interaction causes the molecules to be attracted to the surface, the energy of attraction being:
or
where a is the polarizability of the adsorbed molecule, C,,, is an effective surface charge, z is the distance the molecule is from the surface, and E$, which is usually a strong function of z, is the electric field intensity normal to the surface. In going from the gaseous to the adsorbed state, a molecule undergoes changes in its degrees of freedom, The minimum change occurs when a transla262
ANALYTICAL CHEMISTRY
tion in the gas phase becomes, on adsorption, a vibration normal to the surface. At low temperatures, a free rotation in the gas phase can become hindered on the surface. In extreme cases (very low temperatures) the gas on the surface can form an immobile layer in which all gaseous translational degrees become vibrational degrees and all rotations become hindered rotations or librations on the surface. This extreme case is never found under the conditions normally employed in GSC in which there is a t most a change in two degrees of freedom. King and Benson, by calculating the field intensity over the (100) plane of an idealized All03 surface, showed that adsorbents like Aln03 do possess strong surface electric fields. They calculated
King and Benson, through a detailed treatment, were able to calculate separation factors and heats of adsorption for the hydrogen isotopes which agreed quantitatively with experimental values. They ascribed the barrier to rotation, necessary to explain the separations of the ortho- and paraspecies, to the difference in ( ~ 1 1 and C U ~ the parallel and perpendicular components of the polarizability of hydrogen. Their interpretation differs from that of Mortensen and Eyring (19) who, by assuming that non-polar van der Waals' forces were responsible for the adsorption, attributed the barrier to a greater orientation and tighter packing of the orthohydrogen on the surface. They based their conclusion on the statement that orthohydrogen in the liquid has a lower molar volume than parahydrogen which would cause it to be closely packed and more restricted. However, Lambert (11) found that ortho- and para-hydrogen differ by less than 0.1% in their liquid molar volumes. This slight difference could not account for the barrier of approximately 0.7 kcal. per mole which is needed to explain the separations. It is unlikely that any type of non-polar van der Waals' interactions could account for such a large barrier. If, in fact, the barrier is caused by electrostatic forces, then graphite, on which Sandler (20) was able to separate orthoand parahydrogen, must have an ionic surface layer. This would account for the observation that water forms a monolayer on graphite. Thus, the electrostatic model not only accounts for the interaction energies between the molecules and the surface, it also predicts what the preferred orientation of the molecules on the surface will be. There exists both parallel and perperdicular components of the surface electric fields. However, in the electrostatic model, the adsorbed molecules are assumed to form a mobile monolayer in which their motion parallel to the surface is not affected by the parallel electric field. On the other hand, the perpendicular electric field causes the preferred orientation
,
to be that in which the electric field normal to the surface acts in the direction of greatest polarizabiity. For diatomic molecules, this direction is along the internuclear axis APPLICATION OF THEORY
The electrostatic theory finds wide applications in GSC. It is able to predict the order of elution of gases from many adsorption columns. In the simple case of the rare gases on an A120acolumn, the gases elude from the column according to their relative polarizabilities. Also, as predicted by the theory, a plot of the logarithm of their retention times as a function of their polarizabilities leads to essentially a straight line as shown in Figure 1. Methane was included with the rare gases since its polarizability is isotropic. The fact that it behaves like a rare gas and elutes from the column between krypton and xenon is further support for the theory. It is somewhat surprising that none of the gases deviated significantly from the line in Figure 1 since the interaction is also a function of the sizes of the molecules. In fact, z in Equation 1 can be replaced by (r-TO) where T is the distance between the adsorbed molecule and the surface and TO is the radius of the adsorbed molecule. Additional support for the electrostatic theory comes from the work of Gant and Yang (6) who, in separating the isotopic methanes on a charcoal column (-3.5’ to 150’), found that the retent8iontimes descreased in accordance with the decrease in polarizabilities as deuterium or tritium was substituted for the hydrogen in methane. In the case of molecules whose polarizabilities and sizes are essentially the same-i .e., the hydrogen isotopesother factors, such as the vibrational and rotational energies on the suface, govern the separations. An interesting example of this effect is the chromatographic separability of oxygen and argon. It is well known that these species are difficult to separate on an adsorption column, especially a t room temperature or higher. The electrostatic theory offers a plausible explanation. Not only are the two similar in size but also defined the average polarizablity of 02, as l/a(all 2 a i ) , is 1.60 and the polarizability of argon is 1.63 (12). These values are so close that there is little difference in the interaction energies of the molecules with the surface. The vibrational energies on the surface are also very close since they are functions of the square root of the total mass (10). Therefore, to separate the gases some other parameters must be changed. One choice is to change the rotation of the oxygen on the surface by lowering the temperature. At the lower temperature the rotation can become hindered and the effective polarizability
+
Table II. Temperature Dependence of the Retention of Gases on Linde 5A Molecular Sieve Column ( I )
T ( ” K.) 313 329 352 373
1000/T 3.19 3.04 2.84 2.68
HZ
Nz
0 2
1 1 1 1
3.5 3.0 2.5 1.5
12.0 8.0 5.5 3.0
becomes, not the average, but the polarizability along the internuclear axis. It is then not surprising that oxygen and argon can be separated on a number of adsorbents a t low temperatures (8, 14). The electrostatic theory also explains the fact that polar molecules or molecules with large quadrupole moments are irreversibly adsorbed on chromatographic columns. Since the dipole- or quadrupole-electric field interactions are much stronger than the induced dipole-electric fields intereactions, the molecules are held more strongly to the surface. In order for these molecules to be eluted, the highly active sites on the surface must be blocked either by impurities or chemically, as has apparently been done by Scott (21). Perhaps the most striking application of the electrostatic theory is to the adsorption of gases on molecular sieves. The prevailing belief is that the adsorption and separation of gases on molecular sieves are governed by the sizes of the “holes” or “channels” in the molecular sieve structure. This idea apparently originated with McBain (15) who coined the phrase “molecular sieve” for the zeolites after observing that chabazite, a naturally occurring zeolite, rapidly adsorbed the vapors of water, methyl and ethyl alcohol, and formic acid while those of acetone, ether, and benzene were largely excluded. McBain’s hypothesis has been vigorously expounded by Barrer (2) who has made exhaustive studies of gas adsorption on the zeolites. However, the fact that symmetrical chromatographic peaks are obtained with molecular sieve columns tends to repudiate the hypothesis that separations on the column occur because of some “sieve” action. If molecules must diffuse through 4 to 13 A. diameter “holes” or “channels,” then one would expect to see a significant amount of tailing in the peaks. The absence of tailing suggests that the process is not diffusion controlled, and that adsorption on molecular sieve columns is similar to adsorption on other chromatographic columns. In spite of all the investigations there is still general confusion as to why molecules separate on molecular sieves, and which molecules are adsorbed by a particular sieve. For example, Hersh (7) indicates that while silica gel
CH4 19.5 13.0 9.0 6.0
i-Butane
Ethane
24.5 11.0 4.5 1.5
40.0 14.0
...
...
and activated alumina adsorb n-butane and isobutane, Linde synthetic Molecular Sieve 5.4 adsorbs n-butane but excludes isobutane on the basis of size. Such a result would suggest that the mechanism of adsorption on molecular sieves was different from that on the other adsorbents. However, Spencer (22) found that isobutane is adsorbed in a Linde 5A chromatographic column and its retention on the column shows a strong temperature dependence. The fact that ortho- and parahydrogen are separated in a molecular sieve column ( I ) suggests that adsorption on molecular sieves is similar to adsorption on other adsorbents used in GSC in that electostatic interactions play a dominant role. The separation factor for ortho- and parahydrogen on Linde 5.4 is 1.40 which approximates that found on an A1203 column (1.22) (18) and a SiOz column (1.39) (17). A detailed analysis of the separation Nz, 02, CH4 and isobutane on a of Hz, Linde 5A Xolecular Sieve column (22) supports the contention that adsorption on sieves is similar to adsorption on other adsorbents. In Table I1 is listed the relative retentions, based on HI, of the gases as a function of temperature. These values were taken from the chromatograms of Spencer (22) which showed the Hzpeak to change little with temperature. The reversal in the elution time of oxygen and nitrogen on the 5A molecular sieve used by Spencer (compare Table I and Figure 11) suggest that this sieve possesses unusual adsorptive properties. We unsuccessfully attempted to duplicate the experiments of Spencer using Linde 5A molecular sieve which had been activated a t 450’ C for 24 hours. In our experiments there was no reversal of oxygen and nitrogen and isobutane was not adsorbed on the column a t any temperature. A plot of the change in relative retention with temperature produces the curves shown in Figure 2. The heats of adsorption listed in the Figure were calculated using the relationship:
TiTz
Q
AT
ti
AH = R - l n -
(3)
where tl and t2 are the relative times a t temperatures T1 and Tz. Equation 3 is obtained from the classical equation of Martin and Synge (16). VOL. 38, NO. 2, FEBRUARY 1966
263
~~
Table 111.
Nz 0 2
i-C4Hio
Differences in Adsorption Entropies and Heats of Adsorption
-2.70 1.77 -0.20 3.0 5.06 9.10
352 373 352 373 352 373
-4.55 7.50 0.43 9.45 15.6 27.1
2.7
-. 2 .o -3.0
all the gases, except methane, occurs a t the same temperature. As in the case of adsorption on A1203, methane with its isotropic polarizability is uneffected by changes in electric field strength and has essentially the same heat of adsorption over the total temperature range. The similar behavior of nitrogen, oxygen, and isobutane tends to refute the contention of Spencer that isobutane
The curves in Figure 2 show the type of behavior which is predictable from the electrostatic theory. At the higher temperatures adsorption occurs predominantly a t the most active site but, as the temperature is lowered, these sites become blocked by irreversibly adsorbed gases. The less active sites must then participate in the process. Within the limitations of the data it appears that irreversible adsorption of
adsorbs on the “external” surface of the sieve while the other gases adsorb on the “internal” surface. In fact, there is little or no direct experimental evidence to suggest that it is any but the ‘‘external” surface area which is effective in adsorption by the sieves. The magnitude of the heats of adsorption is further indication that electrostatic forces are responsible for the adsorption. This is also supported by the changes in the entropies of the gases which are listed in Table 111. Because of the ideal behavior of methane the entropies of adsorption are calculated using methane as a reference, that is, in Table I11
A(AS)
=
AS, - A ~ c H ~
and
(4)
A(AH) = AH, - AHcH~
m;;:;; 1
where 2 refers to S 2 ,O2 or i-C4H10. The adsorption entropies for the high and low temperature regions of the curves in Figure 2 are listed in Table I11 along with the corresponding heats of adsorption. The vaporization entropies are included for comparison. The large entropy changes over a 20” C. span in temperature suggests that adsorption does occur over different sites on the surface, and that the more active sites are saturated a t the lower temperatures. The differences in the entropies of adsorption and vaporization is further evidence that different forces are responsible for the two processes. Additional support for the hypothesis that electrostatic forces govern adsorption on molecular sieves comes from the investigations of Barrer and Gibbous (3). They found that the heat of adsorption of NHI on Linde sieve X is a function of the substituted cation in the sieve, the heat value increasing as the polarizing power of the cation increases. This is the expected behavior if Equations l or 2 govern the adsorption process.
..........
I...
E
---
i
E
I SO BUTANE
CONCLUSION
0“r
Kcal/rnole ARE’SHOWN ON THE CURVES
ii
I
2.7
II
2.8
II
I
I
2.9
3.0
3.1
IO00
I I
3.2
T
Figure 2. Relative gas retentions on Linde Molecular Sieve 5A as a function of temperature
264
ANALYTICAL CHEMISTRY
The electrostatic theory affords interpretation of much of the data of GSC. The examples cited deal with simple molecules on a limited number of adsorbents. The method can and will be extended to more complicated systems, such as the separation of large polar organic molecules. It has been the inability to separate such molecules on adsorption columns which has limited the applicability of GSC. From the knowledge of gas-solid interactions as supplied by the electrostatic theory it should be possible to synthesize new adsorbents which will extend the versatility and utilization of gas-solid chromatography.
LITERATURE CITED
( 1 ) Bachmann, L., Bechtold, E., Cremer, E., J . Catalysis 1, 113 (1949). (2) Barrer, R. M., DiscussiOns Faraday SOC.7, 135 (1962). (3) Barrer, R. M., Gibbous, R. &I., Trans. Faraday SOC.59, 2569 (1963). (4) Basila, M. R., J . Phys. Chem. 11, 2223 (1962). (5) Brunauer, S., Emmett, P. H., Teller, E., J . Am. Chem. SOC.60,309 (1938). (6) Gant, P. L., Yang, K., Ibid., 85, 5063 (1964). (7) Hersh, C. K., “Molecular Sieves,” Chap. 4, Reinhold, New York, 1961. (8) Jones, K., Halford, P., Nature 202, 1003 (1964).
(9) King, J., Jr., J . Phys. Chem. 67, 1397 (1963). (10) King, J., Jr., Benson, S. W., J . Chem. Phys., 44, in press. (11) Lambert, M., Phys. Rev. Letters 4, 555 (1960). (12) Landolt Bornstein, “Zahlenwerte
-
und Funktionen aus Physik, Chemie, Astronomic, Geophysik, und Technik, 6th ed., Vql. 1, 509-512, SpringerVerlag, Berlin, 1951. (13) Langmuir, I., Colloid Symp. Monogr.
3, 72 (1925). (14) Lard, E. E., Horn, R. C., ANAL. CHEM.32, 878 (1960). (15) McBain, J. W., Colloid Symp. Monogr. 4, 7 (1926). (16) Martin, A. J. P., Synge, R. L. M., Biochem. J . 35, 1358 (1941).
(17) Mohnke, M., Saffert, W., “Proc. 4th International Gas Chromatography Symposium,” M. Van Swaay, Ed., Butterworth, Washington, D. C., 1962. (18) Moore, W. R., Ward, H. R., J . Am. 80, 2909 (1958). Chem. SOC. (19) Mortensen, E. A., Eyring, J., J . Phys. Chem. 64,433 (1960). (20) Sandler, Y . L., Ibid., 58, 58 (1954). (21) Scott, C., “Proc. 4th International Gas Chromatography Symposium,” bl. Van Swaay, Ed., p. 36, Butterworth, Washington, D. C., 1962. (22) Spencer, C. F., J . Chromatog. 1 1 , 108 (1963).
RECEIVED for review August 4, 1965. Accepted October 11, 1965.
Formamide as a Stationary Phase in Gas Chromatography JOSEF NOVAK and JAROSLAV JANAK Institute of Instrumental Analytical Chemistry, Czechoslovak Academy o f Sciences, Brno, Czechoslovakia
b Formamide is the most polar known stationary phase for gas chromatography. It has a substantially lower volati‘ity than N,N-dimethylformamide or acetylacetone and produces a very low signal in the flame ionization detector. One mole of formamide gives a signal equivalent to 0.3 gram atom of paraffinic carbon. It is suited for the separation of highly volatile polar substances. The retention indices according to Kovbts, corresponding to the polar functional groups, are two to three times higher than on other known polar phases. The peaks of strongly polar substances possessing small moleculessuch as methanol, acetone, etc., are symmetrical even with untreated supports which have relatively high surface areas-e.g, Sterchamol, crude firebrick, aluminum silicates, etc. The values of the C term of the Van Deemter equation are half to one fifth of those obtained for phases commonly used. Formamide is well suited as a stationary phase when capillary columns are used, as well as in trace analysis. The use of formamide as a stationary liquid for the gas chromatographic analysis of complicated mixtures of oxygen-, SUIfur-, nitrogen-, or halogen-containing compounds, is demonstrated by chromatograms.
Keulemans, Kwantes, and Zaal (10) have recommended dimethylformamide as an extremely polar stationary phase for gas chromatography. Retention data of some hydrocarbons on formamide a t 0” C. were reported by Podbielniak and Preston (17). Later, diethylformamide (15) and diphenylformamide (22) were recommended for separation of both saturated and unsaturated hydrocarbons. The use of dimethylformamide is confined to temperatures of about 0’ C. (4, IO),and to detectors of low or medium sensitivity-e.g., a katharometer. The vapor pressure of dimethylformamide is about 9 mm. of Hg (9, 18) a t 40’ C. I n textbooks on gas chromatography, dimethylfonnamide is referred to as a typical phase of extremely high polarity, but in practical work it is rarely used because of its high vapor pressure. These properties of dimethylfonnamide probably were the reason for neglecting the possibility of utilizing unsubstituted formamide, with the exception (17) referred t o above. Thus, other important recently described properties of amides (6, 14) have so far escaped attention. Consideration of the properties of amides induced us to reopen the investigation of formamide as a stationary phase for gas chromatography.
T
Apparatus. Measurements were carried out on Chrom 11, a chromatograph of Czechoslovak origin (Laboratory Instruments, Prague), using flame ionization detection. This instrument is intended for work with either packed or capillary columns (16).
HE lower carboxylic acid amides are strongly associated and exhibit anomalous polarity and vapor pressure (9). These properties make them an interesting subject for chromatogaphic studies.
EXPERIMENTAL
The experiments were devised to furnish information on some of the abovementioned aspects of exploiting formamide as a stationary phase in gas chromatography. The packed columns were stainless steel tubes, 85 cm. long and 6 mm. in i.d.; the capillary columns were 50 meters long and 0.2 mm. in i.d. The V , values were measured a t 10” intervals over a range from 30” to 70” C. For practical applications, the working range was from 30” to 40” C. The injection Dort was maintained a t 100” C: throughout. Substances Used in Column Preparation and Analysis. For the preparation of packed columns four types of support were used, including one of high activity. The supports were chosen with regard t o their typical properties, as revealed by electron microscopic examination ( 8 ) . Celite 545 (Johns-Manville Co., Ltd., London, England), particle size 0.170 to 0.149 mm., specific surface area 1 to 2 sq. meters per gram. Crude firebrick (Calofrig, N.E., Borovany, Czechoslovakia), particle size 0.25 to 0.35 mm., specific surface area 10 to 15 sq. meters per gram; raw material for production of Rysorb supports (8)’ S t e r c h a m o 1 (St e r c h a m o 1we rk e G.m.b.H., Dortmund, West Germany), particle size 0.25 to 0.35 mm., specific surface area 7 to 8 sq. meters per gram. Xlusial (Slovnaft, N.E., Bratislava, Czechoslovakia), a technical product used as an inorganic ion exchanger, substantially a sodium aluminum silicate containing 8.5% &03,specific surface area 120 to 160 sq. meters per gram. Alusil is more polar than silica gel or aluminum oxide ( 7 ) . Formamide (VEB Laborchemie, Apolda, East Germany) was applied in the amount of 25% (w./w.) directly, VOL. 38,
NO. 2, FEBRUARY 1966
265