Process for steam reforming of oxygenates and catalysts for use in the process

Abstract

In a process for steam reforming of oxygenates, especially at low steam-to-carbon (S/C) ratios, a feed gas containing oxygenates, such as ethanol, is converted into syngas over a ternary carbide catalyst. Then the reformed gas is either transformed into desired chemicals or mixed into the feed stream to the reformer in a plant, such as an ammonia or methanol plant. The preferred ternary carbide is nickel zinc carbide.

Claims

1. A process for steam reforming of oxygenates, in which a feed gas containing oxygenates, is converted into syngas according to the reactions
C.sub.2H.sub.5OH+3H.sub.2O.Math.2CO.sub.2+6H.sub.2 or
C.sub.2H.sub.5OH+H.sub.2O.Math.CH.sub.4+CO.sub.2+H.sub.2 over a nickel zinc carbide catalyst, and which after use shows nickel solely in the form of alloys, where the syngas is subsequently either transformed into desired chemicals or mixed into a feed stream to a reformer in a plant selected from the group consisting of an ammonia plant and a methanol plant.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) FIG. 1 shows the lattice parameter for face centered cubic (Ni) solid solution in the Ni—Zn system according to the above-cited Stadelmaier and Huettel reference,

(2) FIG. 2 shows the thermodynamic activities for solid solutions of Ni and Zn at 1100 K according to the same reference,

(3) FIG. 3 illustrates the measured exit gas composition in run No. 1 with respect to methanol, ethanol, methane, carbon monoxide and hydrogen as a function of time. Inlet conditions: 4.5 vol % methanol and 4.5 vol % ethanol in nitrogen; total flow 100 Nml/min; temperature 500° C.,

(4) FIG. 4 illustrates the measured exit gas composition in run No. 2 with respect to ethanol, methane, carbon monoxide and hydrogen as a function of time. Inlet conditions: 3.27 vol % ethanol and 9.82 vol % water in nitrogen (S/C=1.5); total flow 100 Nml/min,

(5) FIG. 5 illustrates the equilibrium (eq) and measured exit gas composition in run No. 2 with respect to methane, carbon monoxide and hydrogen as a function of temperature. Inlet conditions: 3.27 vol % ethanol and 9.82 vol % water in nitrogen (S/C=1.5),

(6) FIG. 6 and FIG. 7 show the propanol reforming in run No. 5; exit gas composition as measured with respect to methane, carbon monoxide and hydrogen as a function of time. Inlet conditions: 1400 ppmv methanol, 2.05 vol % propanol and 10.21 vol % water in nitrogen (S/C=1.63),

(7) FIG. 8 illustrates the ethanol reforming in run No. 5; exit gas composition as measured with respect to ethanol, methane, carbon monoxide and hydrogen as a function of time. Inlet conditions: 1300 ppmv methanol, 3.24 vol % ethanol and 9.72 vol % water in nitrogen (S/C=1.47). Total flow: 100 Nml/min,

(8) FIG. 9 illustrates the ethanol and acetic acid reforming in run No. 5; exit gas composition as measured with respect to ethanol, methane, carbon monoxide and hydrogen as a function of time. Inlet conditions: 2100 ppmv methanol, 1.99 vol % ethanol, 1.98 vol % acetic acid and 9.94 vol % water in nitrogen (S/C=1.22). Total flow: 100 Nml/min, and

(9) FIG. 10 illustrates the ethanol glycol reforming in run No. 5; exit gas composition as measured with respect to methane, carbon monoxide and hydrogen as a function of time. Inlet conditions: 1900 ppmv methanol, 1.99 vol % ethanol, 2.13 vol % ethylene glycol and 4.26 vol % water in nitrogen (S/C=0.95). Total flow: 100 Nml/min.

(10) The experiments were conducted in TPX-3A using a U-tube quartz reactor with an inner diameter of 4 mm. Each test was carried out using a catalyst loading of 600 mg (sieve fraction: 300-600 μm), fixed between two layers of quartz wool. The reactor temperature was measured by a thermocouple secured on the outside of the reactor.

(11) Prior to any experiment, the given catalyst was reduced in 200 Nml/min of 10 vol % hydrogen in nitrogen, either with or without 5 vol % methanol. The exact reduction procedure used in each experiment is given in Table 1 below.

(12) During experiments, an oxygenate/water mixture was fed to an evaporator (⅛″ steel tube filled with SiC) by a syringe pump and mixed with nitrogen—typically to a total gas flow of 100 Nml/min. The specific oxygenates and steam-to-carbon (S/C) ratios used in each experiment are also listed in Table 1.

(13) TABLE-US-00001 TABLE 1 Overview of oxygenate reforming experiments Run Oxygenate(s) No. Catalyst Pre-treatment (S/C) 1 12 wt % Ni, 16 Heating to 600° C. at Methanol + wt % Zn/Al.sub.2O.sub.3 5° C./min in 10 vol % ethanol (0) H.sub.2/N.sub.2 (200 Nml/min). Treatment at 600° C. for 6 h in 10 vol % H.sub.2/N.sub.2 (200 Nml/min). 2 12 wt % Ni, 16 Reactor loading from Ethanol wt % Zn/Al.sub.2O.sub.3 run No. 1 (1.50) 3 12 wt % Ni, 16 Treatment at 600° C. for Propanol wt % Zn/Al.sub.2O.sub.3 6 h in 5 vol % methanol, (1.67) 10 vol % H.sub.2/N.sub.2 (200 Nml/min). 4 18 wt % Treatment at 400° C. for Propanol Zn/R-67 6 h in 5 vol % methanol, (1.63) 10 vol % H.sub.2/N.sub.2 (200 Nml/min). 5 6 wt % Heating to 250° C. at Propanol Ni/SK-501 5° C./min in 10 vol % (1.63) H.sub.2/N.sub.2 (200 Nml/min). Ethanol Treatment at 250° C. for (1.47) 0.75 h in 5 vol % methanol, acetic acid + 10 vol % H.sub.2/N.sub.2 ethanol (200 Nml/min). (1.22) ethylene glycol (0.95).

(14) Spent catalysts from the oxygenate reforming tests have been characterized by X-ray diffraction spectroscopy. Except for the first experiment, the S/C corresponds to an oxygen-to-carbon ratio of about 2. The exit gas composition with respect to hydrogen, carbon monoxide, methane, ethylene and ethane was measured by gas chromatography (GC).

(15) Results and Discussion

(16) Table 2 below shows the phase distribution in the spent catalysts as measured by X-ray diffraction spectroscopy. No nickel zinc carbides are found in the spent samples from run Nos. 2 and 3, which both originated from the same sample. In the sample from run No. 2, zinc is only present as ZnAl.sub.2O.sub.4, while zinc is likely present in the spinel phase reported for spent catalyst from run No. 3. These results indicate that the zinc oxide, initially present at the surface of the catalyst, has been consumed by the alumina support during the experiments, e.g. facilitated by the steam in the reaction gas. Graphite (or traces thereof) is found in both samples. The sample from run No. 4 contains Ni.sub.3ZnC, Ni, ZnO and MgO. It is likely that some of the ZnO, impregnated onto the surface of the nickel catalyst R-67, has exchanged place with MgO in the spinel. For the spent catalyst from run No. 5, two nickel alloy phases with lattice constants of 3.618 and 3.568, respectively, have been found. Stadelmaier and Huettel (cited above), who first reported the existence of nickel zinc carbide, found a Ni.sub.3ZnC.sub.07 phase with a lattice constant of 3.65 kX (approximately equal to 3.65 Å) and a second phase with a lattice parameter of 3.53-3.55 kX depending on the zinc content. The latter phase was accredited to a solid solution of Ni and Zn. The formation of a binary Ni—Zn phase (with a lattice constant of 3.568 Å) during run No. 5 is, however, highly unlikely. Such an alloy would have a zinc content of about 20 at % according to FIG. 1. From FIG. 2 it appears that the activity of Zn (at 1100 K) in such a system is quite low. Hence it is doubtful that ZnO has been reduced to form this alloy, either during the reduction step (temperature 250° C.) or during the experiment (max. temperature 600° C., S/C=1.63).

(17) The X-ray diffraction spectroscopy results indicate that a spinel carrier, preferably ZnAl.sub.2O.sub.4, should be used rather than Al.sub.2O.sub.3 in order to ensure formation of the nickel zinc carbide.

(18) TABLE-US-00002 TABLE 2 Results from X-ray diffraction (XRD) analysis Run No. XRD Rietveld refinement 2 8.7 wt % Ni, 14.9 wt % graphite, 32.1 wt % γ-Al.sub.2O.sub.3 and 44.3 wt % ZnAl.sub.2O.sub.4 (a = 8.042 Å). 3 (Uncertain Rietveld refinement) 0.7 wt % Ni, 3.1 wt % α-Al.sub.2O.sub.3 and 96.3 wt % spinel (a = 8.012 Å). Traces of graphite cannot be excluded. 4 7.4 wt % Ni.sub.3ZnC (a = 3.652 Å), 3.6 wt % Ni, 8.9 wt % ZnO, 8.8 wt % MgO, 4.4 wt % α-Al.sub.2O.sub.3 and 66.9 wt % spinel (a = 8.081 Å). 5 0.5 wt % Ni-alloy 1 (a = 3.618 Å) and 11.9 wt % Ni- alloy 2 (a = 3.568 Å) plus 8.6 wt % ZnO, 19.7 wt % graphite and 59.3 wt % spinel (a = 8.094 Å).

(19) FIG. 3 shows the exit concentration of methanol, ethanol, methane, carbon monoxide and hydrogen, measured during run No. 1 using a total gas flow of 100 Nml/min, containing 4.5 vol % methanol and 4.5 vol % ethanol in nitrogen, at a reactor temperature of 500° C. It appears from the Figure that the alcohols are completely converted at 500° C., and the production of methane, carbon monoxide and hydrogen is rather stable over the measurement period. No ethylene or ethane was formed. The Figure cannot account for all the carbon in the feed. While carbon dioxide was also produced during the experiment (but not quantified), the remaining carbon cannot solely have been in the form of carbon dioxide, as the oxygen content in the feed was too low. Hence, it is possible that solid carbon species have formed on the catalyst during the experiment. This is further indicated by the fact that a significant amount of graphite (14.9 wt %) was found by XRD in the spent catalyst from run No. 2, in which the catalyst from run No. 1 was reused.

(20) FIG. 4 shows the exit concentration of ethanol, methane, carbon monoxide and hydrogen, measured during ethanol reforming in run No. 2 at 400-550° C. A total feed flow of 100 Nml/min containing 3.27 vol % ethanol and 9.82 vol % water vapor (corresponding to an S/C of 1.5) in nitrogen was used. As can be seen from the Figure, ethanol is completely converted at 400° C. As expected, the amount of methane produced decreases with temperature, while the amounts of carbon monoxide and hydrogen increase. In fact, there is an excellent agreement between the composition of the exit gas and the equilibrium composition, as shown in FIG. 5.

(21) A fresh loading of the Ni—Zn/Al.sub.2O.sub.3 catalyst was used for propanol reforming in run No. 3. A total gas flow of 100 Nml/min containing 2.07 vol % propanol and 10.35 vol % water (S/C=1.67) in nitrogen was used. While complete propanol conversion was observed at 400° C., the pressure drop over the reactor increased rapidly, and so the experiment was aborted. The failure of this catalyst is most likely due to zinc being consumed by the alumina support rather than forming a ternary carbide with nickel, e.g. facilitated by the somewhat higher water content compared to that of run No. 2.

(22) In run No. 4, a Zn/R-67 catalyst was used for propanol reforming. In this experiment, the feed consisted of 1400 ppmv methanol, 2.07 vol % propanol and 10.18 vol % water (S/C=1.63) in nitrogen. Again, the experiment was aborted because of an increasing pressure drop over the reactor. While a nickel zinc carbide was identified in the spent catalyst by X-ray diffraction, some free nickel (which could facilitate carbon formation) was also found in this sample (see Table 2).

(23) Using a Ni/SK-501 catalyst, propanol reforming was accomplished in run No. 5. In this case the feed gas consisted of 1400 ppmv methanol, 2.05 vol % propanol and 10.21 vol % water (S/C=1.63) in nitrogen. A total gas flow of 100-150 Nml/min was used. FIG. 6 and FIG. 7 show the resulting exit gas composition with respect to methane, ethylene, ethane, carbon monoxide and hydrogen, measured at temperatures between 400 and 600° C. As seen from the Figures, the primary products are carbon monoxide and hydrogen (and most likely also carbon dioxide). The production of carbon monoxide and hydrogen increases with the temperature, just as expected. At all conditions, very low amounts of hydrocarbons are produced. While the gas chromatography equipment used was not calibrated to quantify propanol, the chromatograms confirm complete conversion of propanol at 400° C.

(24) In FIG. 8, the results from ethanol reforming at 400° C. over the same catalyst loading are shown. The feed consisted of 1300 ppmv methanol, 3.24 vol % ethanol and 9.72 vol % water (S/C=1.47) in nitrogen, and the total flow was 100 Nml/min. As can be seen from the Figure, ethanol is converted completely into methane, carbon monoxide and hydrogen at 400° C. Only insignificant amounts of ethylene and ethane are produced.

(25) Steam reforming of a mixture of ethanol and acetic acid was also attempted in run No. 5, using a feed flow of 100 Nml/min consisting of 1900 ppmv ethanol, 1.98 vol % acetic acid and 9.94 vol % water (S/C=1.22) in nitrogen. The resulting exit gas composition with respect to ethanol, methane, ethylene, ethane, carbon monoxide and hydrogen, measured at temperatures between 400 and 500° C., is shown in FIG. 9. Again, the main products are methane, carbon monoxide and hydrogen (as well as carbon dioxide). Slight decreasing trends of the three species can be observed at each temperature step, which could cause catalyst deactivation. Apart from a few measurements, the Figure shows full conversion of ethanol. Unfortunately, acetic acid could not be properly detected by gas chromatography, hence its degree of conversion is unknown. Unidentified peaks were present in the chromatograms.

(26) Finally, steam reforming of ethylene glycol was carried out at 400 and 500° C., likewise during run No. 5. A gas mixture of 1900 ppmv methanol, 2.13 vol % ethylene glycol and 4.26 vol % water (S/C=0.95) was fed to the reactor at a flow of 100 Nml/min. FIG. 10 shows the composition of the exit gas with respect to methane, ethylene, ethane, carbon monoxide and hydrogen. At 400° C., only hydrogen can be quantified, except for the first data point where some carbon monoxide was measured. As during the reforming of ethanol and acetic acid, the hydrogen production decreases over time. At 500° C., some carbon monoxide and a small amount of methane are also produced. The degree of conversion of ethylene glycol is unknown, because it could not be properly detected by gas chromatography. Unidentified peaks were present in the chromatograms.

(27) In conclusion, steam reforming of ethanol, propanol, acetic acid and ethylene glycol, respectively, has been carried out over a range of Ni—Zn based catalysts. A fixed oxygen-to-carbon ratio of 2 was used.

(28) Reforming of ethanol was successfully carried out over a 12 wt % Ni-16 wt % Zn/Al.sub.2O.sub.3 catalyst, reaching gas compositions close to equilibrium. Reforming of propanol over the same catalyst, however, caused a rapid increase in the pressure drop over the reactor. XRD on spent catalyst indicated formation of ZnAl.sub.2O.sub.4 rather than nickel zinc carbide. Steam reforming of propanol over an 18 wt % Zn/R-67 catalyst also resulted in clogging of the reactor. In this sample, both Ni.sub.3ZnC and free nickel could be identified by XRD. Both propanol and ethanol reforming could successfully be carried out over a 6 wt % Ni/SK-501 catalyst. Steam reforming of an ethanol/acetic acid mixture as well as of ethylene glycol over the same catalyst also turned out to be possible, although with some signs of deactivation. The spent catalyst showed nickel solely in the form of alloys.

(29) The formation of nickel zinc carbide is crucial in order to avoid carbon formation at low steam-to-carbon ratios. This requires the use of a support having no (further) affinity towards spinel formation with zinc. An obvious choice in this respect is ZnAl.sub.2O.sub.4.