290
Energy & Fuels 1991,5, 290-299
Role of Microporous Surface Area in Uncatalyzed Carbon Gasification R. H. Hurt,* A. F. Sarofim, and J. P. Longwell Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 Received May 18, 1990. Revised Manuscript Received November 19, 1990
The gasification in oxygen and carbon dioxide of two high-surface-area synthetic carbons was studied in order to investigate the role of microporous surface area in uncatalyzed carbon gasification. The identification of the appropriate total surface area for carbon gasification is critical to the calculation of meaningful intrinsic reactivities and is the logical starting point in the quest for an improved fundamental understanding of the factors that determine carbon gasification reactivity. Two microporous chars having markedly different macropore volumes were selected for study. Low-temperature gasification reactivities, vapor adsorption isotherms, and adsorption equilibration times were measured for the chars at various stages of conversion, and individual surface features of char particles were examined as a function of conversion by a captive-particle SEM technique. Gasification of the highly macroporous Spherocarb carbon is believed to be strictly intrinsic between 400 and 600 "C, with complete reactant penetration into microporous regions lying between larger pores. There is evidence that the rate of gasification of the nonmacroporous sucrose char, on the other hand, was limited by slow microporous diffusion causing incomplete penetration of carbon dioxide into the micropores. The severity of microporous diffusion limitations during the gasification of most carbons should lie between that observed for the highly macroporous Spherocarb carbon, with its highly accessible microporous area, and the essentially nonmacroporous sucrose char with its less accessible microporous area.
Introduction Most of the surface area of high-surface-area carbons lies within micropores, which have diameters less than 20 A. There is, at present, much uncertainty surrounding the role of micropores in gasification and thus much uncertainty in the identification of the appropriate total surface area. The identification of the appropriate total surface area for carbon gasification is important in the development of gasification or pore structure models and in the calculation of meaningful intrinsic reactivities. A properly measured intrinsic rate would be a true property of the carbon surface and could be fundamentally related to (or at least correlated with) other properties of the carbon surface. For this reason, the identification of the appropriate total surface area is the logical starting point in the quest for an improved fundamental understanding of the factors that determine carbon gasification reactivity. It is well-known that carbon surface areas measured by vapor adsorption techniques are often a strong function of temperature and the choice of adsorbate.' For example, very large discrepancies between areas measured by carbon dioxide and nitrogen are common and Nandi et a1.2 have presented evidence that such differences are associated with the occurrence of restricted diffusion in carbon micropores. Restricted diffusion occurs when the size of the pore approximates the size of the diffusion molecules. The diffusing molecules, then, are at all times under the influence of the potential energy field associated with the adjacent pore walls, and the rate of diffusion is slow, activated, and very sensitive to the size of the pores and the size, shape, and kinetic energy of the diffusing species.' The nitrogen surface area of many coals and carbons is low because slow diffusion at the temperature of nitrogen ad-
* Author to whom correspondence should be addressed. Current address: Combustion Research Facilitv. Sandia National Laboratories, 8361, Livermore, CA 94550. 0887-0624/91/2505-0290$02.50/0
sorption measurements (77 K)prevents significant nitrogen penetration into very small micropores, having openings of approximate diameter 3.5 A or smaller, during the time allotted for adsorption equilibration? Several reviews are available on the topic of diffusion in fine pores.'p3 The conditions under which restricted diffusion controls, and the microporous surface area does not fully participate in, gasification are not well established. Laurendeau4 questions the accessibility of measured total surface areas in his review of gasification kinetics, while restricted diffusion was expected to be important only during low conversion for the gasification of an anthracite char in oxygen.5 Johnson6 cites evidence for micropore widening during gasification of coal chars, while Rist et al.' observed a characteristic micropore diameter of 20 A, independent of conversion. Duttas observes that the rates of gasification of his chars are proportional to the surface area lying in pores with diameters larger than 30 A, while there is evidence for micropore widening and thus microporous reaction during the gasification of an anthracite char.g Further, Gavalas'O expects that slow diffusion in micropores will confine the reaction front to a region in the vicinity of the large pores, and thus formulates his model (1) Walker, P. L. Jr.; Austin, L. G.; Nandi, S. P. In Chemistry and Physics of Carbon; Walker, Jr. P. L., Ed.; Marcel Dekker: 1966; Vol. 2, pp 257-371. (2) Nandi, S. P.; Walker, P. L. Jr. Fuel 1964, 43, 385. (3) Ruthven, D. M. Principles of Adsorption and Adsorption Processes; Wiley: New York, 1958; Chapter 5, p 124. (4) Laurendeau, N. M. Prog. Energy Combust. Sci. 1978,4, 221. (5) Walker, P. L. Jr. Fuel 1980,59, 809.
( 6 )Johnson, J. L. Prepn. Pap-Am. Chem. Soc., Diu. Fuel Chem. 1975,20 (4), 85.
(7) Rist, L. P.; Harrison, D. P. Fuel 1985.64, 291. (8)Duttta, S.; Wen, C. Y.; Belt, R. J. Ind. Ena. Chem. Process Des. Deu. 1977, 16, 1. (9) Kawahata, M.; Walker, P. L. Jr. Proceedings of the Fifth Carbon Conference; Pergamon Press: New York, 1963; Vol. 2, p 251. (10)Gavalas, G. R. AIChE J 1980,26, 4, 577.
0 1991 American Chemical Society
Uncatalyzed Carbon Gasification
explicitly considering only the larger pores, assuming that the micropores contribute to the reaction an amount that is proportional to the large pore surface. Simons," on the other hand, considers reaction in all of the pores measured by carbon dioxide adsorption. Finally, several studies have demonstrated that carbon reactivity is not well correlated with measured total surface a r e a ~ . ~ J ~ J ~ There is, clearly, much uncertainty in the literature surround this topic, as well as a lack of experimental information. In this study, we attempt to identify the appropriate total surface area for the uncatalyzed gasification by investigating two carbonswith high microporous surface areas, but varying degrees of macroporosity. One of the carbons has an extensive macropore network, which divides the carbon into smaller microporous regions, whereas the second carbon is uniformly microporous. Low-temperature gasification rates were measured for various particle sizes of the two chars. The surfaces of the partially reacted chars were examined by a captive-particle SEM technique, and their pore structures were investigated by measuring vapor adsorption isotherms (for C02 and N,) and adsorption equilibration times. The results of this experimental study apply to low temperatures where chemical kinetics are traditionally presumed to be dominant, conditions which will be shown to be more difficult to define than previously thought. Materials a n d Experimental Procedures Spherocarb is a pure, high-surfacearea synthetic carbon in the form of spherical particles (60/80 mesh) sold by Analabs, which contains an extensive macropore network as is clear from SEM micrographs and from the measured pore size distribution for Spherocarb supplied by Niksa." A second carbon was made by pyrolyzing reagent grade sucrose to yield a high-area material, which is, however, essentially nonmacroporous, having almost no measurable pore volume in pores larger than 100 A.18 The sucrose char was prepared from a low-temperature char supplied by Dr. J. K. Floessls by heat treatment in nitrogen in a small tube furnace, in a reproducible manner, involving a slow temperature ramp followed by a 1-h holding period at lo00 OC. Spherocarb was used as recieved. Experiments were performed on smaller particle sizes of both Spherocarband summe carbon. Spherocarbparticles, 60/Wmesh (177-250 pm), were crushed with a small mortar and pestle and sieved to -w -e- - - O-0
a
h
0.8 O.g:
0.01
0
7-
I
I
I
I
I
I
1
2
3
4
5
6
time
7
Isec)
Figure 10. Rate of carbon dioxide uptake into initially evacuated Spherocarb at 273 K.
dioxide a t the low relative pressures used in the Dubinin theory apparently does not result in the filling of the large (20 A diameter) micropores. This discrepancy between nitrogen and carbon dioxide areas for highly converted chars has been observed elsewhere and played an important role in the original recommendation of carbon dioxide as an adsorbent for the measurement of coal and char micropore area as distinguished from micropore The nitrogen area is, then, in certain situations, unsuitable for the normalization of gasification rates, but a comparison of its value to the value of the carbon dioxide area may provide some insight into the average micropore size at high conversion, as follows. Pore volume filling at low relative pressures is though to be due to overlap, in very fine pores, of the potential energy fields associated (25) Lamond, T. G.; Marsh, H. Carbon 1964, I , 281.
with the adjacent pore surfaces.26 Enhanced adsorption can also be the result of a "cooperative effect" in larger micropores with diameters up to about 20 This may provide an upper limit on a characteristic pore size present in sucrose carbon at high conversions. A lower limit may be identified by considering the values of the two apparent surface areas at high conversion. The difference between the amount adsorbed corresponding to monolayer coverage ( n = Apom/Aadaorbate) and that corresponding to complete volume filling ( n = V-/ Vhbte) is insignificant for pores with diameters below about 10 A. (Monolayer coverage leaves no room for pore in these pores (dpore< 3dadaorba4 filling.) The factor of about 2 difference between the two areas measured here would not be expected to occur until the pore diameter was about 20 A, which serves then as an estimate of the characteristic micropore diameter for highly reacted sucrose carbon. The appearance of this effect at high conversion is, again, suggestive of micropore widening and gasification within micropores for sucrose carbon.27 Quantitative Estimation of the Effectiveness Factor for Sucrose Carbon Gasification. The observed effect of particle size on gasification reactivity of sucrose carbon in Figure 7 is a classical indication of incomplete reactant penetration. The interesting feature of the data in Figure 7 is that the effect of particle size is rather small between particle diameters of 180 and 50 mm, but increases substantially with further diameter reduction. This behavior is characteristic of the gasification of a carbon in which there are diffusion limitations within microporous domains that are smaller than the particle itself. Grinding of such a particle would cause only modest increases in reactivity, until the point was reached where significant fracture of the microporous grains began to occur. We undertook to formulate the simplest possible model that incorporates the feature of diffusion-limited microporous regions, in order to illustrate and quantify this concept. The microporous volumes between macropores are of irregular and unknown shape. For simplicity, a sucrose carbon particle is modeled as a collection of accessible but internally diffusion limited microporous grains, which are spherical and have monodisperse distribution of radii. Access to the microporous grains is provided by larger pores, which, in the case of sucrose carbon, may be (26) Gregg, S. J.; Sing, K. S. W. Adsorption, Surface Area and Porosity, 2nd ed.; Academic Preas: London, 1982. (27) Evidence for microporous gasification in the case of Spherocarb will be discussed in a later section.
Uncatalyzed Carbon Gasification
0
251
A
Energy & Fuels, Vol. 5, No. 2, 1991 297
Measured values
..... Microporous 'grain' mdel
..A
reaction rate a t 20%conversion in i atm. C O at ~ 830 OC
0.0
0.1
0.2
0.3
0.4
0.5
(pJ Figure 11. Effect of particle size on sucrose char gasification reactivity: comparison of measured values and prediction of microporous grain model. l/Radius
the relatively few macropores18 and/or some fraction of the wider micro- and mesoporosity. The observed gasification rate Robsof such a particle of radius Rp is a function of the intrinsic reaction rate Ri, the effective microporous diffusivity D,, and the grain radius R , which comprise the three parameters in this model. The observed rate is given by Roba = qRi, where q is the effectiveness factor calculated from Ri, R,, ' and D, according to the classical formulation. We need now to consider the effect of grinding on the grain size R, It is assumed that the particles fracture along random planes during grinding and not preferentially along grain boundaries or near the particle surface. After fracture there will be a distribution of both particle and grain sizes and shapes. The calculation of the exact particle and grain sizes and shape, along with the solution to the reaction/diffusion problem for fragments of spheres, is very difficult and not appropriate within the framework of this very simple model. This grinding effect can be easily modeled by assuming that both the fractured particles and grains are spherical and monodisperse. The radii of the fractured grains can be calculated as follows: The external area of the initial grains is given by
is quite consistent with the model of accessible but internally diffusion limited microporous grains. The leastcm2/s, Ri = 0.026 squares parameters are D, = 2.7 X lom7 mi&, and R , , = 27 pm. The corresponding grain area or large-pore area is 0.083 m2/g. The diffusivity is between 2 and 3 orders of ma nitude lower than the effective Knudsen diffusivity in 10-% pores with a tortuosity factor of 5 and a porosity of 0.2, which is 1.0 X lo4 cm2/s (the mean pore diameter for sucrose carbon at 0% conversion is 12.5 A). Tortuosity factors as high as 20 may exist in carbons,28which may account for some of the difference. Even if the tortuosity factor were 20, the observed diffusivity would be significantly lower than the Knudsen diffusivity and, therefore, in the restricted diffusion range. Gasification, by widening pores and increasing diffusivities, would be expected to increase accessibility and reduce or eliminate the particle size effect over the course of conversion. The particle size effect is, indeed, largest at low conversions, but significant reactivity differences persist throughout conversion. The persistence may be a result of the phenomenon of gasification-inducedcarbon densification, which does occur during sucrose gasification,l9 and which would, to some extent, offset the pore widening accompanying gasification. Alternative interpretations of the sucrose particle size effect have been considered. The increase in reactivity with decreasinggarticle size cannot be attributed to an increase in carbon dioxide surface area upon grinding, as can be seen in Table 11. Pronounced inhibition of carbon dioxide gasification by carbon monoxide has been reported and can cause a particle size effect if carbon monoxide produced during gasification accumulates in the particle interior.29 It can be difficult, because of this effect, to obtain uniform activation of carbons in carbon dioxide without the addition of carbon monoxide to the reactant gas to negate the effect of the small CO buildup in the particle interior.30 Since the sucrose particle size effect is the same during gasification in a 50150 carbon dioxide/carbon monoxide mixture, the observed effect of particle size during carbon dioxide gasification is not associated with product inhibition during otherwise kinetically limited gasification. Neither can segregation of the carbon into fractions having different reactivities be responsible for the effect, as the particle size effect was still observed when the small particles were obtained from larger particles while retaining all of the material. Tidjani et alS3lhave observed structural changes in
The external area of grains after fracture is Ag,ext
= Ag,ert(before) -k
AAp,ext
where AApCxtis the external particle area created during grinding and is given by hAp,ext
= ( 3 / P g ) ( l / R p - l/Rp(before))(l - 01)
where €I1 is the porosity in large pores, or between the grains. el is assumed to be small, which is quite a good assumption for sucrose carbon. This equation is an expression of the fact that all new external particle surface area generated by fracture is also new grain external area. The new grain radius can now be calculated from the first expression above. Figure 11 is a plot of the gasification reactivity at 20% conversion taken from Figure 7 for the four particle sizes, along with the predictions of the model generated using the optimum values of the parameters determined by nonlinear least-squares regression. The particle size effect
(28) Hutcheon, J. M.; Longstaff, B. Roc. Ind. Carbon Graphite Conf., London 1958,259. (29) Rand, B.; Marsh, H. Carbon 1971,9,79. (30) Austin, J. G.; Walker, P. L. Jr. AIChE J. 1963, 9, 303. (31) Tidjani, M.; Lachter, J.; Kabre, T. S.;Bragg, R. H. Carbon 1986, 24, 4,447. (32) Floe"* has measured activation energies for the carbon dioxide gasification of a similar sucrose carbon, which ale0 exhibited a particle size effect. His smaller particles exhibited a slightly (15%)higher activation energy than the larger particles, a trend which is consistent with diffusion limitations, and the reverse of that expected if the smaller particles were more reactive due to an effect of grinding on carbon chemistry. This result indicates that the reactant penetration is at least to some extent incomplete. (33) L2/D was, in practice, calculated by fitting the data to an approxunate solution of the transient diffusion equation valid up to t = O.l6(LyD),at which point adsorption is 87% complete, as in Walker et al.' or, or short characteristic diffusion times, from the time required for 87% completion of adsorption.' The solution used is for unsteady diffusion into a sphere at constant external pressure following an initial step change in presaure. The adsorption experimenta were carried out at constant volume, instead of constant pressure, but changes in the external pressure during adsorption were relatively small for the small samples used and were neglected.
298 Energy & Fuels, Vol. 5, No. 2, 1991
graphite using X-ray diffraction during intensive grinding in both a ceramic and a steel ball mill. The particle size effect for sucrose char is not believed to be due to an effect of grinding on structure, because grinding of Spherocarb in the same manner resulted in essentially no change in reactivity. The outstanding difference between the two carbons is pore structure. Quantitative Estimation of the Effectiveness Factor for Spherocarb Carbon Gasification. The reactivity of Spherocarb carbon, unlike that of sucrose carbon, is essentially independent of particle size. This is a necessary but not sufficient condition for intrinsic gasification, due to the possibility of incomplete reactant penetration into microporous grains that are much smaller than the particle itself. The absence of a particle size effect indicates at least that the microporous regions are quite small, making microporous diffusion limitations unlikely since the severity of diffusion limitations scales as L2/D. The following analysis provides support for notion that the Spherocarb gasification is intrinsic and that the difference in the diffusion length scales for sucrose and Spherocarb carbon is sufficient to explain their different gasification behavior. We assume for this analysis that the diffusivity in Spherocarb micropores is equal to the difcm2/s) determined fusivity in sucrose micropores (2.7 by the microporous grain model. (This diffusivity is for carbon dioxide and is expected, therefore, to be too low. Oxygen has a smaller kinetic diameter than CO2= and has been observed to diffuse faster than nitrogen in 4 8,molecular sieve, while nitrogen diffuses faster than carbon dioxide above room temperature in carbons2.) For this analysis one needs to know, in addition, the length scale for microporous diffusion in Spherocarb carbon, or the characteristic radius of microporous grains, which we define here as those regions containing no pores larger than 100 A in diameter. With this definition, the grain radius can be calculated knowin the amount of surface area lying in pores larger than 100 , a quantity which is obtainable
!i
(34) There are two ways of interpreting the diffusivity in the value of Lz/D obtained by fitting the uptake curve to the solution of the transient diffusion equation. First, if the appropriate driving force for fine pore diffusion in a given case is the total concentration, then the diffusivity obtained is the true diffusivity. If, however, the appropriate driving force is the gas-phase concentration only, with the adsorbed species not diffusing, the diffusivity obtained is less than, but related to, the true diffusivity.' One doee not, in general, known what the appropriate driving force for diffusion in a given system is, and must, therefore, recognize the possibility that the true diffusivity is larger than the diffusivity obtained directly from the uptake curve. Second, a finite rate of dissipation of heat generated upon adsorption may often hinder further adsorption and lengthen the observed rate of uptake.% Mass-transfer limitations within the apparatus, and to and through the bed of particles can also limit the rate of uptake. The observed rate of uptake of carbon dioxide into Spherocarb is, in fact, quite comparableto the rate of helium flow through the sample stopcock and into the sample bottle. The equilibration times measured here may reflect to a large extent various mass-transfer limitations external to the particle. Third, some restricted diffusion coefficients have been observed to be pressure or concentration dependent,* increasing with decreasing pressure. Since the extrapolation from adsorption conditions to reaction conditions involves a large reduction in concentration in the porea, the measured and extrapolated diffusivity may be too low. Fourth, the diffusivitiea were extrapolated by using the temperature dependence of the Knudsen diffusion coefficient, which should represent a lower limit on the temperature sensitivity of the diffusivity. Finally, the microporous diffusivities used are those for carbon dioxide diffusion, and can be expected to be lower than those for oxygen, which has a d e r kinetic diameter." Oxygen has, in fact, been observed to diffuse faster than nitrogen in 4 molecular sieve, while nitrogen diffuses faster than carbon dioxide above room temperature in carbonsz.) In each case the approximations involved are such, that the effectiveness factor calculated from the measured and extrapolated diffusivity should be regarded as a minimum effectivenew factor, defining an upper limit on the severity of diffusion limitations during Spherocarb gasification. (35) Sircar, S. Carbon 1981,19,153. (36) Sevenster, P. C. Fuel 1969,38 403.
6
Hurt et al. from mercury porosimetry data for both Spherocarb17and sucrose carbons.ls Microporous grains according to this definition have an external area of 0.092 m2/g and a characteristic radius of 24 pm for sucrose carbon, and an external area of 15 m2/gm and a radius of 0.15 pm for Spherocarb. The large difference reflects the limited macroporosity of sucrose carbon in contrast to the extensive macroporosity of Spherocarb. Effectiveness factors for Spherocarb oxidation, calculated by using this length scale and the microporous diffusivity from the sucrose model, are unity for each of the measured reaction rates in Figure 6 between 400 and 600 "C. The different gasification behaviors of sucrose and Spherocarb can therefore be explained in terms of the differing length scales only. It is interesting to note that this analysis predicts an onset of microporous diffusion limitations at higher temperatures in the gasification of Spherocarb (see Figure 6). Interpretation of the Adsorption Equilibration Times, Characteristic times for equilibration of carbon dioxide adsorption were measured for both chars. The measured amounts of carbon dioxide adsorbed as a function of time were fit to the solution of the transient diffusion equation in a sphere to obtained L2/D under adsorption condition^.^^ The adsorption of carbon dioxide in Spherocarb at both 0 and 45% conversion was complete after several seconds. The adsorption into 45% converted Spherocarb was 87% complete in 0.95 s, yielding a characteristic diffusion time (L2/D)of 5.9 s. For unreacted sucrose carbon, L2/Dfor 180pm diameter particles was 6.1 X lo4 s (or 16.9 h), while at 20% conversion only 17.8 s. The microporous regions within sucrose carbon particles are clearly much more resistant to penetration by carbon dioxide at adsorption conditions than are the microporous regions within Spherocarb particles. Although a reliable extrapolation of diffusion rates to reaction temperatures is not possible, a estimation of the effectiveness factor for the two different carbons may provide insight into their gasification behavior. The Thiele modulus for first-order reaction in a sphere is given by 9 = [ ( R 2 / D 8 ) / ( 1 / k , ) ] 1 /where 2 , R 2 / D is the characteristic diffusion time (estimated from adsorption measurements), 0 is the porosity, and k, is the intrinsic rate constant per unit volume. l/kv is given by l/kv = [C,MW,/np IF, where C, is the concentration of the reaction gas, M$, is the molecular weight of carbon, n is the moles of reacting gas consumed per mole of carbon, ppis the particle density, and r is the reciprocal of the intrinsic reaction rate with units of g/(s g). The analysis yields an effectives factor of 0.92 for the Spherocarb reaction at 400 "C. The assumptions made in this calculation are such that 0.92 should be regarded as a minimum effectiveness This analysis provides additional evidence that the gasification of sucrose carbon is not significantly influenced by restricted diffusion limitations. Values of the effectiveness factor of 1 X lo-' for unreacted sucrose and 0.45 for 20% converted sucrose were obtained. The very low effectiveness factor at 0% conversion reflects that carbon's very long adsorption equilibration time. This very rough analysis predicts little or no diffusion limitations during the gasification of Spherocarb and strong microporous diffusion limitations for sucrose carbon, which disappear over the course of conversion. These predictions are in agreement with the general trends observed in this study.
Conclusions There is much evidence for gasification within the mi-
Energy & Fuels 1991,5, 299-303 cropores of the synthesis carbons. Further, several quantitative analyses of reaction and diffusion during Spherocarb oxidation indicate that reactant penetration is complete and that the reaction rate is strictly kinetically limited. The accessible total surface area for Spherocarb is expected to be the 640 m2/g measured by carbon dioxide adsorption, and the intrinsic kinetics for oxidation of Spherocarb in 0.21 atm of O2 are given by rate (g/(s m2) = 3.0 X lo3e*IRT, where R is in kcal/(gmol K). The value 36 kcal/mol is a typical activation energy for a low-temperature char or carbon but is substantially lower than activation energies measured for oxidation of some graphites,'l the difference presumably arising from the different degrees of purity and ~rystallinity.'~This level of understanding of diffusional processes is necessary for the fundamental treatment of many aspects of carbon gasification, including a pore structure modeling effort recently undertaken in this lab~ratory.'~ There is evidence for sucrose carbon, on the other hand, that reactant penetration is incomplete. The nature of the effect of particle size on sucrose char reactivity is consistent with restricted diffusion limitations in microporous grains with radii of 27 pm. The severity of diffusion limitations scales as L2/Dand depends, therefore, on the local microporous diffusivity,
299
a parameter determined by micropore size, and the characeristic size of the microporous regions, a parameter determined by macroporosity. The major difference between the behavior of the two synthetic carbons in this study is thought to arise from a difference in the characteristic size of microporous grains in the two carbons. Hippo et aL3' have also emphasized the importance of feeder pores to determine the likelihood or severity of microporous diffusion limitations. The severity of microporous diffusion limitations for most chars should lie between that for the highly macroporous Spherocarb and that for the essentially nonmacroporous sucrose carbon. It is therefore expected that the carbon dioxide surface area will be the appropriate total surface area for the gasification of many pure carbons, although restricted diffusion limitations may be important during low-temperature gasification of chars with little or no macroporosity.
Acknowledgment. Support of this research by Exxon Research and Engineering Co. is gratefully acknowledged. This paper is based on work supported also under a National Science Foundation Graduate Fellowship. (37) Hippo, E.; Walker, P. L. Jr. Fuel 1975,54 245.
A Simple Method for Analysis of Nitrogen and Phenolic Compounds in Synthetic Crude Naphtha Tadashi Yoshida,**+Pierre D. Chantal, and Henry Sawatzky Energy Research Laboratories, CANMET, 555 Booth Street, Ottawa K 1 A OG1, Canada Received June 5, 1990. Revised Manuscript Received October 9, 1990
A simple and rapid analytical method for trace amounts of nitrogen and phenolic compounds in hydroprocessed Boscan naphtha has been developed by combined use of column adsorption chromatography with basic alumina and GC technique. Based on the breakthrough curves of total nitrogen and total sulfur on basic alumina, an optimal condition was chosen for separating selectively nitrogen and phenolic compounds from the naphtha. The nitrogen compound fraction was further subdivided into groups of pyridine, pyrrole, aniline, and indole types on GC column. The main components in the nitrogen and phenolic compound fractions were identified by gas chromatography-mass spectrometry and determined by gas chromatography. The amount of nitrogen compound fraction in the naphtha was 2.0 wt '% and pyrrole type was most abundant. The amount of phenolic compound fraction was less than 0.4 wt '% and oxygen compounds other than phenols were not found.
Introduction The synthetic crude naphthas derived from coal, heavy oil, tar sand bitumen, etc. generally contain trace amounts of polar compounds. These materials often bring about the deposition of insoluble sediments and g u m s during the storage of naphtha'9 ahd are also harmful environmentally when naphtha is used. Therefore, these polar compounds should be removed from naphtha and the development of a new analytical method will be required for the charac'Present address: Government Industrial Development Laboratory,Hokkaido 2-17-2-1Tsukiiu-Higashi, Toyohira, Sapporo 004, Japan.
Table I. Ultimate Analysis of Boscan Naphtha ultimate analvsis. w t % naphtha Boscan
C
H
S
N
84.7
14.0
0.97
0.35
0 0.05
* 0.01
H/C 1.97
terization of polar compounds present in synthetic crude naphthas. Although many studies on the separation and characterization of polar compounds from petroleum?" coal(1) Cooney, J. V.; Beal, E. J.; Hazlett, R. N. R e p r . Pap.-Am. Chem. SOC.,Diu. Fuel Chem. 1984,29(3), 247-255. (2) Mayo, F. R.; Lan,B. Y.Ind. Eng. Chem. R o d . Res. Dev. 1986,25, 333-348.
0887-0624/91/2505-0299$02.50/00 1991 American Chemical Society
