Ind. Eng. Chem. Res. 1992,31, 2251-2255 Johnson, J.; Chiantelb, A. J.; Cashost, H. W. Application of Light Scattering to Diesel Fuel Stability Problems. Znd. Eng. Chem. 1966,47(61,1226-1230. Kap, W. M.; Crawford, M.F. Heat Transfer: Laminar Flow Inside Tubes. In Conuectiue Heat and Mass Transfer;McGraw-Hill: New York, NY,1980,pp W 9 7 . Kendall, D. R.; Mille, S. S. Thermal Stability of Aviation Kerosines: Techniques to Characterize Their Oxidation Properties. Znd. Eng. Chem. Prod. Res. Deu. 1986,25(2),360-365. Lee, S. C.; Tien, C. L. Optical Constants of Soot in Hydrocarbon Flames. Symp. (Znt.) Cornbut. [Proc.] 1981,lath, 1159-1166. Melles Griot. Melles Griot Optics Guide 3; Mellea Griot: Irvine, CA, 1985; pp 5656. Mille, S. S.; Kendall, D. R. The Quantification and Improvement of the Thermal Stability of Aviation Turbine Fuel. J. Eng. Gas hLrbines Power 1986,108,381-386. Mori, Y.; Futagami, K. Forced Corrective Heat Transfer in Uniformly H e a t 4 Horizontal Tubes. Znt. J. Heat Mass Transfer 1967,10,1801-1813. Roguemore, W. M.; Pearce, J. A.; Harrison, W. E.; Krazinski, J. L.; Vanka,S. P. Fouling in Jet Fuels: New Approach. Prepr.-Am. Chem. SOC.,Diu. Pet. Chem. 1989,34,841-849. Szetela, E. J.; Giovanetti, A. J.; Cohen, S. Fuel Deposit Character-
2261
istics of Low Velocity. Trans. ASME 1986,108,460-464. Taylor, W. F. Kinetics of Deposit Formation from Hydrocarbons. Znd. Eng. Chem. Prod. Res. Deu. 1969,8 (2),375-380. Taylor, W. F. Deposit Formation from Deoxygenated Hydrocarbons: I. General Featuree.. Znd. Eng. Chem. Prod. Res. Deu. 1974,13 (2),133-138. Taylor, W. F. Deposit Formation from Deoxygenated Hydrocarbons: 11.. Effect of Trace Sulfur Compounds. Znd. Eng. Chem. Prod. Res. Deu. 1976,15 (11, 64-68. Taylor, W. F.; Wallace, T. J. Kinetics of Deposit Formation from Hydrocarbon Fuels at High Temperatures. Znd. Eng. Chem. Prod. Res. Deu. 1967,6 (4),258-762. Taylor, W. F.; Wallace, T. J. Kinetics of Deposit Formation from Hydrocarbons: Effects of Trace Sulfur Compounds. Znd. Eng. Chem. Prod. Res. Deu. 1968,7 (3), 198-202. van de Hulst, H. C. Scattering and Extinction Experiments as a Tool. In Light Scattering by Small Particles; Dover: New York, NY; Chapter 18. Weast, R. C. Handbook of Chemistry and Physics, 50th ed.; Chemical Rubber Co.: Cleveland, OH, 1970;pp C-75-C-538.
Received for review January 28, 1992 Accepted June 19,1992
RESEARCH NOTES Organophosphorus Compounds as Coke Inhibitors during Naphtha Pyrolysis. Effect of Benzyl Diethyl Phosphite and Triphenylphosphine Sulfide Significant reduction in the rate of coke formation during naphtha pyrolysis was achieved by adding benzyl diethyl phosphite or triphenylphosphine sulfide to the feed. Although the yield of carbon oxides was reduced, there was no effect of these additives on the hydrocarbon yields. Addition of these organophosphorus compounds significantly reduced the concentration of metals, such as iron, nickel, and chromium, incorporated in the coke. A previously proposed model for coke inhibition due to the formation of a passivating metal-phosphorus complex could satisfactorily correlate the
data. A major problem encountered during naphtha pyrolysis is the deposition of coke on the reactor walls. This coke deposit reduces the overall heat-transfer coefficient and increaaee the pressure drop across the reactor, necessitating periodic shutdowns for decoking. Frequent decoking operations result in loes of production, affect the coil life, and increase fuel and utility costs. Several techniques have been attempted to reduce coke formation during pyrolysis, and Vaish and Kunzru (1989) have summarized the available information. Recent studies have shown that phosphorus compounds can inhibit the coking rates without affecting the product yields (Naberezhnova et al., 1983; Boone, 1983). Ghosh and Kunzru (1988)achieved a significant reduction in the rate of coke formation by adding triethyl phosphite to the feed, whereas Vaish and Kunzru (1989)found that triphenyl phosphite was a more effective coke inhibitor than triethyl phosphite. The objective of this study was to investigate the effectiveness of benzyl diethyl phosphite and triphenylphosphine sulfide (a compound containing both phosphorus and sulfur) as coke inhibitors during naphtha pyrolysis. In addition, the coke morphology and changes in the metal content of the deposited coke due to addition of benzyl diethyl phosphite have been studied. Experimental Section The pyrolysis experiments were conducted in a jetstirred 316 stainless s-1 reactor (length 0.031 m; volume
19.1 X 10-6 m3). To measure the amount of coke deposited, a small Inconel 600 disk (diameter 7 mm) was suspended in the reactor, periodically weighed outside the reactor, and resuspended. The details of the experimental setup and procedure have been described earlier (Kumar and Kunzru, 1985). For all the runs presented below, coke formation was measured on Inconel 600 disks which had earlier undergone several coking-decoking cycles. Compared to this, the coke formation on a fresh Inconel plate was approximately 20% lower and gradually increased with the number of coking-decoking sequences. After five to six oxidation-reduction cycles, the coking rate was independent of the number of coking-decoking sequences The reproducibility of the data was within 110%. The increase in the coking rate due to repeated coking-decokingis moat likely caused by the increase in surface roughness (Crynes and Crynes, 1987). The feed was either desulfurized or straight-run naphtha of the same boiling range, and the detailed naphtha characteristics are given in Table I. The sulfur content of the straight-run naphtha was 80 ppm whereas that of desulfurized naphtha leas than5 ppm. The phosphorus compounds used were either benzyl diethyl phosphite (BDP) (bp 440-442 K at 25 mm) or triphenylphosphine sulfide (TPPS). The feeding system was modified when TPPS (mp 435-437 K; decomposition temperature 653 K) was used as the additive. To avoid deposition of this additive on the relatively cooler section of the vaporizer inlet, the naphtha was introduced through a capillary at a point in the vaporizer where the temper-
0888-5885/92/2631-2251$03.00/00 1992 American Chemical Society
2262 Ind. Eng. Chem. Res., Vol. 31, No. 9, 1992 Table I. Characteristics of Naphtha Feed (ASTM Distillation) density (298 K) 700 kg/m3 av mol wt 90 vol % distilled" temp, K vol % distilled" temp, K IBP 313 70 388 90 410 10 340 30 361 FBP 433 50 375 a
Phosphorus content,ppn
0 0 A 50 0
x 0 c
.-
5
IBP, initial boiling point; FBP, final boiling point.
100
-
15-
x Y
," u
Phosphorus content
0 A 0
V
10-
1
a
0 100 200 500
5-
OL
0
I
1.0
I
2.0
I
3 .O
I
4 .O
I
5.
Time,h
Figure 2. Effect of triphenylphosphine sulfide on the rate of coke formation (feed, desulfurized naphtha + 2.0% (v/v) acetone; 5" = 1103 K,6 = 0.7, 7 = 0.6 8).
"0
1 .o
2.0
3 .O
4.0
5.0
lime, h
Figure 1. Effect of benzyl diethyl phosphite on the rate of coke formation (feed, desulfurized naphtha; T = 1103 K,6 = 0.7,T = 0.6 8).
ature was above the melting point of TPPS. When BDP was used as the additive, the phosphorus content of the condensed organic or aqueous phase was determined photometrically (Das, 1991).
Results and Discussion Pyrolysis experimenta were conducted with or without additives (BDP or TPPS) at atmospheric pressure in the temperature range of 1073-1103 K. For most of the runs, the weight ratio of steam to naphtha, 6, and the space time, 7 , were kept fixed at 0.7 kg/kg and 0.6 8, respectively. Since TPPS was not easily soluble in either naphtha or water, it was dissolved in acetone and then added to the naphtha feed. Effect of Phosphorus Content of Feed. The presence of BDP in the feed significantly reduced the rate of coke formation (Figure 1). For instance, with naphtha containing 200 ppm phosphorus (as BDP), the asymptotic coking rate, r,, was reduced by 87.7%. The rate of coke formation was initially high and gradually decreased to an asymptotic value in approximately 2.5-3 h. Such a trend of coking rate with run time has been reported by several investigators. Under identical operating conditions, BDP was found to be a more effective coke inhibitor than triphenyl phosphite. The decrease in r, with increasing phosphorus concentration was more appreciable at lower concentrations and then gradually decreased at higher phosphorus concentrations. Some runs were also taken with straight-run naphtha (sulfur content = 80 ppm) as feed. At 1103 K, the asymptotic rate of coke formation was 60% lower than that for desulfurized naphtha. The inhibiting effect of sulfur
due to the formation of passivating metal sulfides is well established (Bajus et al., 1981; Sahu and Kunzru, 1988). In comparison, the reduction in r, in the presence of 80 ppm phosphorus (as BDP) was estimated to be 70%. Addition of 100 ppm phosphorus (as BDP) to the straight-run naphtha resulted in a 62% reduction in r,. Thus, for the range of conditions studied, the inhibiting effect of phosphorus was not as pronounced as for desulfurized naphtha. However, more data are needed on the combined effects of phosphorus and sulfur on the rate of coke formation. To study the effect of sulfur-containing organophosphorus compound on the rate of coke formation during pyrolysis, TPPS was taken as an additive. Since TPPS is not readily soluble in either naphtha or water, ita effect on the coking rate was studied by dissolving the required amount in acetone and then mixing this solution with the feed to give a concentration of 2.0 vol % acetone in the naphtha. Reference runs were also taken for naphtha containing 2.0 vol % acetone and no TPPS. Rates of coke formation with run time at two different phosphorus concentrations are shown in Figure 2. There was no significant effect of acetone on the asymptotic rate of coke formation. At identical conditions (2' = 1103 K, T = 0.6 8, 6 = 0.7), ra for acetone containing naphtha (no TPPS) was 10.0 X kg/(m2.h) compared to 9.8 X kg/(m2.h) for the desulfurized naphtha. At the same phosphorus concentration, TPPS was found to be more effective than BDP in reducing the rate of coke formation. A comparison of Figures 1 and 2 shows that the rate of coke formation at 50 ppm phosphorus concentration using TPPS was comparable to the rate of coke formation at 100 ppm phosphorus concentration using BDP. A possible reason for the higher effectiveness of TPPS could be that, at this reaction temperature, TPPS can dissociate into (c6&,)2p*, and C6HsS'. In radicals such as (C,&)2P+, case the P=S bond is broken, then for each molecule of TPPS two radicals which can passivate the metal walls would be generated, whereas in the case of BDP only one passivating radical is formed per molecule dissociated (refer to reactions A, B, and C). However, more direct evidence is required to conclusively show that both P and S act separately in TPPS. Since TPPS is a solid at room temperature and is not easily soluble in the feed, the use
Ind. Eng. Chem. Res., Vol. 31, No. 9, 1992 2253 Table 11. Effect of Bemy1 Diethyl Phosphite on Gaseous Product Yields (9% Maor) (T= 1103 K,T = 0.6 E, I = 0.7; Feed, Derulfurised Naphtha) naphtha with phosphorusphosphorus component free naphtha 100 ppm 500 ppm 14.78 14.86 14.23 methane 25.97 26.11 26.28 ethylene 2.02 2.03 1.95 ethane acetylene 1.16 1.16 1.19 0.71 0.72 0.77 propane 12.45 12.52 12.79 propylene 0.46 0.47 0.48 isobutane 0.80 0.80 0.82 n-butane 2.97 2.98 3.06 1-butene 4.78 4.80 4.93 2-butene l,&butadiene + isobutene 5.79 5.82 6.02
c4+
4.55
4.58
4.92
total
76.44
76.85
77.44
Table 111. Effect of Temperature, T , and I on r , and Yield of Carbon Oxides (Additive. BDP) yield of carbon oxides, wt phosphorus % feed content of temp, 6, 1@r., fWd.DDm K ka/k T , S kg/(m*.h) CO COz 0 0 0 100 100 100 200 200
500 0 100 200
500 0 100 200
500
1103 1103 1103 1103 1103 1103 1103 1103 1103 1088 1088 1088 1088 1073 1073 1073 1073
0.7 0.7 0.96 0.7 0.7 0.96 0.7 0.96 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7
0.6 0.37 0.6 0.6 0.37 0.6 0.37 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
9.8 8.0 8.7 2.2 1.5 1.7 0.8 0.9 0.8 7.3 1.8 0.9 0.6 6.0 1.2 0.7 0.4
3.73 1.37 5.11 1.71 0.68 1.90 0.47 0.98 0.47 2.57 1.02 0.62 0.21 1.20 0.83 0.60 0.13
4.47 2.13 4.09 2.16 1.5 1.18 0.39 0.82 0.73 2.81 1.23 0.72 0.31 1.43 0.93 0.67 0.20
of TPPS is cumbersome and was not studied further. Product Yields. The major gaseous products were methane, ethylene, propylene, butenes, l,&butadiene, and bobutene. In addition, undesirable CO and C02were also produced in significant amounts. Addition of either BDP or TPPS had no significant effect on the hydrocarbon gaseous product yields (CO, C02 free) and the detailed steady-state product distribution for two different BDP concentrations is shown in Table 11. Although not shown here, there was also no effect of BDP on the product yields when straight-run naphtha was used as the feed. In contrast to this, the yields of CO and C02decreased when either BDP or TPPS was added to the feed, and the effect of BDP on the yield of carbon oxides is given in Table 111. These trends are in good agreement with the results reported by Vaish and Kunzru (19891, who used TPP as the additive, and with the results of Boone (1983), who employed proprietary phosphorus-containing coke inhibitors. The carbon oxides during pyrolyeis are formed mainly at the surface by either steam gasification or steam reforming reactions (Holmen et al., 1982). Metals such as nickel, iron, and chromium present in the coke are known to catalyze the steam reforming reaction. The lower concentration of CO in presence of phosphorus is most probably due to the interaction of the phosphorus compounds with the metal surface thereby reducing the catalytic effect of the metals on the steam reforming reaction.
Table IV. Phosphorus Distribution in Condensed Aqueous and Organic Phases (Additive, Benzyl Diethyl Phosphite; T = 1103 K.I= 0.6 m. d 3: 0.7) phosphorus content of feed, PPm phosphorus content of org phase, ppm phosphorus content of aq phase, ppm % recovery in org phase Irecovery in aq phaee total recovery, %
100 797 38 57 24 81
200 1313 77 63 27 90
500 4125 204 58 26 84
A similar reasoning has been suggested by Froment (1990). Phosphorus Distribution in Condensed Organic and Aqueous Phases. The phosphorus distribution in the aqueous and organic product, when BDP was used as the inhibitor, is shown in Table IV. The phosphorus selectively concentrated in the organic phase. Since the amount of organic liquids was much less than the weight of condensed water, the phosphorus recovery (as a percentage of the total phosphorus added to the feed) in the organic phase varied from 57 to 63%, whereas the phosphorus recovery in the aqueous phase ranged from 24 to 27%. The balance phosphorus could have been either deposited on the coke or exited as some gaseous product. No attempt was made to determine the fate of the remainder of the phosphorus. This method of phosphorus determination could not be used when TPPS was the additive because it is not possible to convert the TPPS to the orthophosphate state by wet oxidation. Effect of Process Variables on the Rate of Coke Formation. It is well established that rates of coke formation during pyrolysis of hydrocarbons depend on the reactor temperature, space time, and steam dilution ratio. The effect of these process variables on r, and yield of carbon oxides with BDP as the inhibitor is given in Table III. For both phosphorus-free and phosphorus-containing naphtha, asymptotic coking rates increased with an increase in temperature or T whereas they decreased with an increase in 6. There was no appreciable effect of temperature on the percentage reduction in the coking rate. For instance, with 200 ppm phosphorus, the percentage reduction in r, at 1073, 1088, and 1103 K was 88.3,87.6, and 87.770, respectively. Coke formation is a complex phenomenon, and several producta such as aromatics, acetylene, and diolefins have been identified as the coke precursors. The concentration of these precursors increases with increasing conversion (large T or high T ) or increasing hydrocarbon partial pressure (small 61,resulting in a corresponding increase in the rate of coke formation. The yield of carbon oxides also increased with an increase in the reactor temperature. Morphology and Metal Content of Deposited Coke. To study the morphology and metal content of coke deposits, a number of runs were taken at the same operating conditions (T= 1103 K, T = 0.6 s,6 = 0.7, run time = 4 h) with varying inlet concentrations of phosphorus. The deposited coke was analyzed using a scanning electron microscope (Jeolco Model JSM-840A) equipped with an energy-dispersive X-ray analyzer (EDAX). Figures 3 and 4 show photomicrographs of coke deposited during pyrolysis of naphtha containing varying amounts of phosphorus. All the photographs are of coke surfaces which were in contact with the gas. In all cases, the coke surface was rough and porous. Major differences in the morphology of coke were not observed due to addition of phosphorus. Both with and without phosphorus, several types of coke morphologies were observed, and Figure 3 shows the nature
2264 Ind. Eng. Chem. Rea., Vol. 31, No. 9,1992
c A.Clobular coke (P=lOO ppm) 10 p m
1-1
8
P = 100 ppm.
1 pm
H
Figure 4. Morphology of filamentous mke obtained during pym lynk of naphtha containingBDP (T= 1103 K,6 = 0.7. I= 0.6 s, run time = 4 h).
B.Globular coke (P=200 ppm)
j -1
10 v m
C.Chaln grovth (P=200 ppm) 10 p m
D.Braided 10 w m
1-1
fllamente ( P = S O O ppm)
1-1
Figum S. Morphology of mke deposits during pyrolpia of naphtha containing BDP (T= 1103 K. 6 = 0.7, r = 0.6 8. m time = 4 h).
of the coke deposited during pyrolysis of naphtha with BDP as the additive. Predominantly, the coke formed wan globular in nature (Figure 3A$). In some cases, the coke globules combined together to form long chains (Figure 3 0 . In addition, braided filamenta (Figure 3D) as well as filamentouscoke was also observed. Such morpholcgiea have also been reported by Marek and Albright (1982) and Albright and Marek (1988). who studied the coke deposition on both polished and unpolished Inconel 800 and aluminized Inconel 800 surfaces. The nature of the filamentous coke is shown in Figure 4. The diameter of the filaments was in the range of 0.7-1.8 pm. As can be seen
in Figure 4, metal particles are located within the body of the filament rather than at the tip. The main conatituenta of these particles were iron and nickel. Such bidirectional f h e n t a have also been reported during the nickel-iron catalyzed decomposition of ethylene (Baker, 1990). An EDAX analysis of the coke surface showed that the main metal constituenta were iron, nickel, and chromium. It is well documented (Baker et al., 1977; Albright and Marek, 1988) that metals such as iron, nickel, and chromium from the reactor walls are incorporated in the coke. These metals, particularly iron and nickel, are known to catalyze coke formation. The concentrations of iron, nickel, and chromium in the coke, as measured from the peak heights of the EDAX plots, were significantlyreduced when either BDP or TPPS was added to the feed. The total metal content, as well as the relative concentrations, were not the same throughout the surface. Estimates of the absolute concentrations of these metals in the coke were made by comparing the peak heights in the actual sample to the peak heights obtained with pure metal samples. The total metal content (Ni + Fe + Cr) of coke formed during pyrolysis of phosphoruefree naphtha varied from 0.7 to 1.0 wt %;the metal was 3040% Fe, 5040% Ni, and 9-12s Cr. In comparison to this, when 500 ppm phosphorus (as BDP) was added to the feed, the total metal content of the coke was reduced to 0.3-0.4 w t %. The metal composition was 65-75% Fe, 13-18% Ni, and 12-20% Cr. These EDAX studies suggest that the organophosphorus compounds inhibit the coking rate by passivating the surface so that the concentration of metala incorporated in the coke is reduced. Modeling of Coke Formation. These d t a show that coke formation is significantly reduced when BDP or TPPS is introduced with the feed. Since the product distribution is not affected, the additive only affecta the heterogeneous coke forming reactions. Most probably, BDP and TPPS passivate the metal surface by decomposing into free radicals which combine with the metal surface. At high temperatures, benzyl diethyl phosphite can decompose in three ways. (CeH&HJPO(OC2H& C,&CHz + (C2Hs0)2P0 (A)
-
(CBH&H~PO(OCZH& (Ce"CHJPO(OC2Hs) + 6c& (B) (C~HsCHJPO(OC2HJz (C6H6CHJP06(0C2HJ + C2Hs* (C)
-
Ind. Eng. Chem. Res., Vol. 31, No. 9, 1992 2255
Acknowledgment The financial support provided by the Department of Science and Technology, New Delhi, for this study is gratefully acknowledged. Nomenclature Kp = equilibrium constant for metal-phosphorus complex formation P = phosphorus concentration in the reactor, weight fraction r, = asymptotic rate of coke formation, kg/(m2.h) rab = asymptotic rate of coke formation on a phosphorus-free surface, kg/ (m2-h) re = asymptotic rate of coke formation for a surface completely covered with metal-phosphorus complex, kg/(m2-h) T = temperature, K Greek Symbols 6 = weight ratio of steam to naphtha, kg/kg 7 = space time, s
4.01
3.0
0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
(rob-’o) x l6:kgl(m2)(h) P
Figure 5. Determination of re and K p at different temperatures.
The phosphorus-containing free radicals thus formed can combine with the metal surface to form a paasivating film. A possible reason for BDP being a more effective coke inhibitor than TPP could be that the metal-phosphorus complex is more stable with BDP or a larger metal surface area is occupied by each molecule of the complex. A model to explain the reduction in coke formation due to the addition of triethyl phosphite or triphenyl phosphite has been presented earlier (Ghosh and Kunzru, 1988,Vaish and Kunzru, 1989). On the basis of this model, the following relationship can be obtained:
where Kp depends only on temperature. On the other hand, fa), - r* depends on the concentration of coke precursors as well as temperature. The concentration of coke precursors are, in general, higher a t higher 7 and lower 6. Therefore, according to eq 1,seta of data at constant 7 and 6 and the same temperature should yield straight lines of the same slope. As shown in Figure 5, the model eq 1could represent the experimental data satisfactorily for all the runs. Compared to the results reported for triethyl phosphite (Ghosh and Kunzru, 1988) and triphenyl phosphite (Vaish and Kunzru, 19891, at the same temperature, rApis lower and Kp higher, which implies that the metal-phosphorus complex is more stable and the rate of coke formation lower on the metal-phosphorus complex formed with benzyl diethyl phosphite. The exothermic heat of adsorption for Kp and the activation energy for ra, as determined from Arrhenius plots, were 100 MJ/kmol and 230.9 MJ/kmol, respectively.
Conclusions This study shows that the rate of coke formation during naphtha pyrolysis can be significantly reduced by adding benzyl diethyl phosphite or triphenylphosphine sulfide to the naphtha. Addition of these compounds does not have any effect on the hydrocarbon yields, but the yield of carbon oxides is reduced. The concentration of metals, such as Fe, Ni, and Cr, incorporated in the coke is significantly reduced due to the addition of these compounds.
Literature Cited Albright, L. F.; Marek, J. C. Coke Formation During Pyrolysis: Roles of Residence Time, Reactor Geometry, and Time of Operation. Znd. Eng. Chem. Res. 1988,27,743-751. Bajm, M.; Vesely, V.; Baxa, J.; Leclercq, P. A.; Rijks, J. A. Steam Cracking of Hydroxycarbona. 5. Effect of Thiophene on Reaction Kinetics and Coking. Ind. Eng. Chem. Prod. Res. Dev. 1981,20, 741-745. Baker, R.T. K. Electron Microscopy Studies of the Catalytic Growth of Carbon Filaments. In Carbon Fibres Filaments and Composites; Figueiredo, J. L., Bemardo, C. A., Baker, R. T. K., HOttinger, K. J., Eds.; NATO AS1 Series E177; Kluwer Academic: Dordrecht, The Netherlands, 1990; pp 405-439. Baker, R. T. K.; Keep, C. W.;Frame, J. A. Origin of Filamentous Carbon Formation from the Reaction of Propane over Nickel. J. Catalysis 1977,47,232-238. Boone, K. Coke Control Extends Furnace On-stream Time. Oil Gas J. 1983,81 (Sept 26), 83-85. Crynes, L.L.;Crynes, B. L. Coke Formation on Polished and Unpolished Incoloy 800 Coupons during Pyrolysis of Light Hydrocarbons. Znd. Eng. Chem. Res. 1987,26,2139-2144. Das, P. K. Benzyl Diethyl Phosphite and Triphenylphosphine Sulfide as Coke Inhibitors during Naphtha Pyrolysis. M.Tech. Thesis, Department of Chemical Engineering, IIT, Kanpur, 1991. Froment, G. F. Coke Formation in the Thermal Cracking of Hydrocarbons, Rev. Chem. Eng. 1990,6,293-328. Ghosh, K. K.; Kunzru, D. Reduction of Coke Formation during Naphtha Pyrolysis Using Triethyl Phosphite. Znd. Eng. Chem. Res. 1988,27,559-565. Holmen, A.; Lindvaag, 0. A.; Trimm, D. L. Surface Effects on the Steam Cracking of Propane. ACS Symp. Ser. 1982,202,45-48. Kumar, P.; Kunzru, D. Kinetics of Coke Deposition in Naphtha Pyrolysis. Can. J. Chem. Eng. 1985,63, 597-604. Marek, J. C.; Albright, L. F. Formation and Removal of Coke Deposited on Stainless Steel and Vycor Surfaces from Acetylene and Ethylene. ACS Symp. Ser. 1982,202,123-150. Naberezhnova, G. N.; Nurtdinov, S. K.; Belyakova, L. D. Organophosphorus Compounds as Inhibitors of Coke Formation during Pyrolysis of Liquid Hydrocarbons. Chem. Abstr. 1983,9!3,25105~. Sahu,D.;Kunzru, D. Effect of Benzene and Thiophene on Rate of Coke Formation during Naphtha Pyrolysis. Can. J. Chem. Eng. 1988,66,80&815. Vaish, S.; Kunzru, D. Triphenyl Phosphite as a Coke Inhibitor During Naphtha Pyrolysis. Znd. Eng. Chem. Res. 1989, 28, 1293-1299.
* To whom correspondence should be addressed. Present address: Department of Chemistry, Delhi Institute of Technology, Delhi, India.
Pradip Das, Surendra Prasad: Deepak Kunzru* Department of Chemical Engineering Indian Institute of Technology Kanpur 208 016,India Received for review March 2, 1992 Accepted April 8,1992