Ind. Eng. Chem. Prod. Res. Dev. 1985, 2 4 , 397-401
It is of practical importance that the tin catalyst has a high activity and can be used repeatedly in the liquefaction of coal. The high activity of the tin catalyst brings in a high yield of liquid products, and repeated use of the catalyst reduces the amount of catalyst usage. In order that coal liquefaction with tin catalyst be employed on a commercial scale, a recovery process for tin must be developed, since the tin catalyst in a powder form is always mixed and diluted with coal ash after usage. A study on the hydrogenation of model compounds typical of coal structure, such as dibenzyl ether and bibenzyl, has been undertaken in order to elucidate a reaction mechanism in the presence of tin oxide catalyst. The results will be published shortly.
MOLTEN Sn
FRAGMENTS
/""V
repolymerization U'
"
COKE
397
sta bil i zat ion
4 hydrogenat ion
1 OIL , ASPHALTENE
Figure 9. Reaction scheme of hydroliquefaction of coal with tin oxide catalyst.
Acknowledgment
conditions. The tin oxide catalyst is reduced to metallic tin, which is present as small lriolten tin particles dispersed in coal slurry under the liquefaction conditions. A reaction scheme of hydroliquefaction of coal with tin oxide catalyst is shown in Figure 9. Fragment radicals formed by thermal cleavage of coal would be stabilized on the molten tin surface. Repolymerization is strongly suppressed by the stabilization effect of molten tin, probably. Almost all fragment radicals formed are hydrogenated to products such as oil and asphaltene. The tin catalyst used twice repeatedly maintained a high activity, indicating that the catalyst was not deactivated significantly. Deposition of carbonaceous materials on the catalyst surface, which is usually the major cause of deactivation of the hydrotreating catalyst, is insignificant in the case of molten tin catalyst. The high activity exhibited by the used tin catalyst results from the fact that the catalyst is not deactivated during the hydroliquefaction reaction.
We are grateful to R. Kaji of Hitachi Research Laboratory for chemical analysis of ash. Registry No. SnOz, 18282-10-5; Fe203, 1309-37-1; Ti02, 13463-67-7;MnO, 1344-43-0;ZnO, 1314-13-2;Co, 7440-48-4;Mo, 7439-98-7; Ni, 7440-02-0. Literature Cited Bertolacinl, R. J.; Gutberlet, L. C.; Kim, D. K.; Robinson, K. K. EPRl AF-547 Project 408-1, 1977. Cochran, S. J.; Hatswell, M.; Jackson, W. R.; Larkins, F. P. Fuel 1982, 6 1 , 831. Han, K. W.; Wen, C. Y. Fuel 1979, 58, 779. Mizumoto, M.; Yamashita, H.; Matsuda. S. Prepr. P a p . Am. Chem. SOC. Div. Fuel Chem. 1983, 28, 218. Neavel, R. C. Fuel 1976, 55, 237. Strong, H. W. British Patent 335 215, 1929. Weller, S.;Pelipetz, M. G.; Friedman, S.;Storch, H. H. Ind. Eng. Chem. 1950, 42, 330. Yarzab, R. F.; Given, P. H.: Spackman, W.; Davis, A. Fuel 1980, 5 9 , 81.
Received for review September 10, 1984 Accepted March 18, 1985
Kinetics of the Vapor-Phase Hydrogenolysis of Methyl Formate over Copper on Silica Catalysts Danlele M. Montl, Mark S. Walnwrlght," and Davld L. Trimm School of Chemical Engineering and Industrial Chemistty, The University of New South Wales, P.O. Box 1. Kensington, New South Wales, 2033 Australla
Noel W. Cant School of Chemistry, Macquarie University, North Ryde, New South Wales, 21 13 Australia
The kinetics of the hydrogenolysisof methyl formate to methanol on ion-exchanged copper on silica catalysts was studied in a recycle reactor at temperatures from 120 to 190 O C and atmospheric pressure. The rate data were fied by a power law model in which the reaction orders with respect to the reactants, methyl formate and hydrogen, were 0.39 and 0, respectively. Carbon monoxide produced by the decarbonylation of methyl formate was shown to inhibit the hydrogenolysis reaction, and a reaction order of - 0 . 1 7 with respect to CO was found. At CO concentrations higher than 1-2 % a continuous loss of catalyst activity was observed.
around 250 "C. Kung (1980) and Klier (1982) have comprehensively reviewed the catalytic and mechanistic aspects of this conventional route to methanol. An alternative two-stage process for the production of methanol involving the carbonylation of methanol to
Introduction
Methanol manufacture on an industrial scale is currently based on the hydrogenation of carbon monoxide over Cu/ZnO catalysts containing alumina or chromia at pressures ranging from 90 to 110 bar and temperatures 0 1 96-432 118511 224-0397$01.50/0
0
1985 American Chemical Society
398
Ind. Eng. Chem. Prod. Res. Dev., Vol. 24, No. 3, 1985
methyl formate and the subsequent hydrogenolysis of this intermediate compound has long been known (Christiansen, 1919). It has recently been claimed (Petrole Informations, 1982) that methanol synthesis using the two-stage process is feasible under comparatively mild conditions (110 OC, 5 bar) in a three-phase reactor. Sorum et al. (1984) have investigated the hydrogenolysis of methyl formate in a slurry reactor over copper-based catalysts. Under their conditions (140-185 "C, 38-100 bar) the reaction rate was first order with respect to methyl formate concentration and zero order in hydrogen pressure with an activation energy of 53.2 kJ/mol. Evans et al. (1983a,b) have studied the hydrogenolysis of various aliphatic formates in the gas phase at atmospheric pressure on a variety of copper-based catalysts. In these studies it was found that the catalytic activity declined when the hydrogenolysis of methyl formate was performed on Raney copper catalysts. The deactivation rate was lower over copper chromite catalysts (Evans et al., 1983~).Stable catalyst activity was reported for the hydrogenolysis of all the other formates investigated. The present paper reports the kinetics of the hydrogenolysis of methyl formate in the gas phase under conditions of stable catalyst activity using a silica-supported copper catalyst prepared by the ion-exchange technique.
Experimental Section Catalysts. The copper catalysts were prepared by ion exchange following the procedure described by Kobayashi et al. (1980). The Aerosil SiOz support materials were obtained from Degussa A.G. The grades used were Aerosil 0x50, 200, and 300, with the numbers referring to the BET surface areas of the supports in m2/g. A 0.21 M copper ammine solution was prepared by the addition of concentrated ammonia to a Cu(N03)zsolution until the pH reached 11. Dry silica support material (15 g) was added to 250 mL of the ammine solution, and the suspension was stirred for 6 h at room temperature while the pH was maintained a t a constant value of 11 by periodical addition of concentrated ammonia solution. The suspension was filtered, and the cake was resuspended in 300 mL of distilled water and stirred for 10 min. The catalyst cake obtained after the second filtration was then transferred to an evaporating basin, dried overnight (16 h) at 110 "C, and calcined at 500 OC for 5 h. The catalysts were crushed, and the fraction with particle diameters between 0.35 and 0.50 mm was taken for testing. The catalysts were carefully reduced in the reactor with hydrogen at 300 "C; initially a 5% H2/Nzgas mixture was used and the reduction was completed in pure hydrogen overnight (flow rates 30 cm3/min-'). Catalyst Characterization. Specific surface areas and pore-size distributions of the samples were determined by nitrogen adsorption-desorption measurements at -196 "C by using a Micromeritics 2100 E surface area analyzer. Copper surface areas were measured by the decomposition of nitrous oxide by using the pulse chromatographic method described by Evans et al. (1983d). The amount of copper in the catalysts was determined by atomic absorption spectroscopy after digestion of the metal with a 3:l HC1:HN03 mixture at 60 "C for 6 h. Kinetic Studies. Apparatus. The reaction rates were determined in a continuous recycle reactor in which the gases were recirculated by a stainless steel bellows pump (Model MB-41, Metal Bellows Corp., Sharon, MA). The gas flow in the recycle loop was 8000 cm3 min-' as measured by a rotameter that was temporarily included in the loop. Since the rotameter itself restricted the flow of gas, it was removed from the recycle loop after the measure-
ment was made. The recycle ratio (defined as the volume of gas recirculated divided by the volume of gas fed to the reactor) was always greater than 20. The reactor was operated at conversion levels between 5 and 50%. These low conversion levels along with recycle ratios in excess of 20 justify the assumption of an ideally mixed continuous reactor. Moreover, at high feed rates the conversions were extremely low and therefore compensated for any reduction in recycle ratio. The recycle loop (0.64-cm 0.d. stainless steel tubing) was heated to 50 OC to prevent condensation of reactants and products. The catalyst (3-4 cm3, 1-1.5 g) was contained in a O.&cm i.d. stainless steel U-tube reactor immersed in a stirred oil bath. The temperature of the reactor was controlled to f l "C. Feed gas flows were maintained constant by the use of calibrated rotameters with needle valves and back-pressure regulators. Methyl formate (Ajax Chemicals Co., 98%, balance methanol) was dosed by passing a primary hydrogen stream through a saturator thermostated at 19.6 "C. The hydrogen was dispersed through a GO sintered glass frit, and the bubbles of ca. 0.1-0.3-cm diameter had a residence of approximately 1 s in the liquid, The saturation temperature was controlled to f O . l K by water recirculated in the jacket of the saturator. The methyl formate partial pressure was calculated on the basis of vapor pressure data available. The degree of saturation was checked by gas chromatography over the range of flow rates used and was found not to depend on the gas flow through the saturator. In some instances the saturator was operated with a methyl formate/methanol mixture to vary the partial pressures of the two components independently. The apparatus was further equipped with gas mixing facilities that enabled diluent gases (Hz, He) as well as CO to be mixed with the primary gas stream. Total inlet flow rates were varied between 100 and 400 cm3/min, and the pressure drop over the catalyst bed was 0.3-0.4 bar. The product stream was sampled by a heated Valco 6-port valve and analyzed with a Gow-Mac Series 550 gas chromatograph fitted with a thermal conductivity detector. Separation of the carbon monoxide (CO), methanol (MeOH), and methyl formate (MF) was performed on a 2 m X 0.32 cm i.d. column of Poropak N that was maintained at 110 OC. Hydrogen was used as carrier gas at a flow rate of 25 cm3/min. Treatment of the Results. Under the conditions applied in this study the major components analyzed in the product stream were methyl formate, methanol, and CO. Traces of C02and HzO were found in concentrations lower than 0.1 % . The C02 was most likely produced by the decomposition of formic acid, which can be formed by hydrolysis of methyl formate. The production of CO by decarbonylation of methanol could be discounted on the basis of earlier findings (Evans et al., 1983b). The following stoichiometric equations describe the reactions occurring in the system CH3OCHO + 2Hz s 2CH3OH (1) CH3OCHO s CH,OH
+ CO
(2)
Conversion (X) and selectivity (S)were calculated from the concentrations of carbon atoms containing products in the reactor exit stream according to the following: (MFli, = [MF] + [CO] + ([MeOH] - [CO])/2 (3) (4)
where brackets indicate concentrations measured by GC
Ind. Eng. Chem. Prod. Res. Dev., Vol. 24, No. 3, 1985
Table I. Catalyst Properties catalvst CuOX50 45 total surface area, m2/g 1.65 copper content, wt % 1.4 copper surface area, m2/g average pore radius, nm 25 bulk density, g/cm3 0.375
Cu200 190 6.6 5.0 13 0.333
Cu300 247 8.2 6.7 10
0.333
analysis and parentheses indicate calculated methyl formate inlet concentration. Methyl formate conversions could be measured with a maximum scatter of f2.5 ?%. These errors were mainly due to the temperature fluctuations in the oil bath (fl"C) and, to a lesser extent, to variations in the flow rates of the different feed gases. The reaction rate r A (mol min-' m-'Cu) was calculated from
by assuming an ideally mixed reactor where F'MF (mol min-') is the inlet molar flow rate of methyl formate and S(m2c,) is the copper surface area of the catalyst in the reactor. The reaction data were fitted to a power law rate expression of the form (7)
where rA is the reaction rate, ko (mol m i d m-2cubar-(E"l)) the preexponential factor, E A (kJ mol-') the activation energy, and Pi(bar) the pressures of the components to the power ai. The dependence of the reaction rate on the reactant partial pressures was measured by fixing the partial pressure of one reactant and measuring the reaction rate at different partial pressures of the other. To maintain constant residence times (space velocities), He gas was used to balance the reactant gas stream to a preset total flow. The influence of the reaction products, CO and methanol, on the reaction rate was assessed separately. In the case of CO, the reaction was run with different amounts of CO added to the reactant gas mixture. The influence of additional methanol in the feed was determined by wing different methanol/methyl formate mixtures in the saturator. Results and Discussion Catalyst Characterization. Some physical properties of the reduced catalysts are summarized in Table I. The mean pore sizes of the catalysts are comparable with the original particle sizes of the silica support materials. The catalyst based on the OX50 support showed the broadest distribution of pore sizes. The catalysts prepared from the high surface area supports (Cu200, Cu300) had markedly narrower pore-size distributions. The pore geometry is therefore mainly determined by the primary particle size and agglomerate characteristics of the catalyst supports. It is interesting to note that the copper loadings and specific copper surface areas are proportional to the support surface areas. Consequently, the copper dispersion does not depend on the type of silica support used. If a silanol site density of 3.2 X 1OI8 m-2 quoted by the manufacturer (Degussa) is assumed, it can be shown that the amount of copper deposited on the supports corresponds to approximately one copper atom per surface silanol group. Catalyst Stability. In some preliminary experiments the long-term behavior of the catalyst was tested over a period of 7 h. At a hydrogen:methyl formate ratio of 3:l
399
and a total gas feed rate of 423 cm3 (STP)min-', constant conversion (61%) and selectivity (96%) were 'found at 187 "C by using 1g of the catalyst Cu200. These conditions corresponded to a gas hourly space velocity (GHSV) of 1.46 X lo4 h-I or a mean residence time in the catalyst bed of 0.25 s. A comparative experiment under the same conditions using 1.2 g of Harshaw 1808 copper chromite catalyst showed a decline in conversion from an initial value of 75% to 62% after 6 h on stream. A t the same time the selectivity was stable at 94%. The stable activity of the ion-exchanged copper on silica catalyst was somewhat surprising in view of an earlier study (Evans et al., 1983~) in which it was found that other copper-based catalysts, and Raney copper catalysts in particular, exhibited a steady decline in their activity for the hydrogenolysis of methyl formate. On the other hand, higher molecular weight formates and acetates could be converted without a loss in activity on the same catalysts (Evans et al., 1984). A similar deactivation has also been reported for the dehydrogenation of methanol to methyl formate [reverse reaction of (111 on Raney copper catalysts (Tonner et al., 1984). Copper well dispersed on an inert support material such as silica, however, did not show any detectible fouling of the catalysts for methanol dehydrogenation. In a previous study of the vapor-phase hydrogenolysis of methyl formate by Evans et al. (1983~)no catalysts were produced by the ion-exchange method. The catalysts prepared by this method have well-dispersed copper crystallites that are much smaller than those in the Raney copper and copper chromite catalysts previously studied. X-ray powder diffraction experiments that were made in order to estimate the copper particle size were unsuccessful and gave evidence that the copper was in a well-dispersed form with particle diameters below 10 nm. The catalysts used in the present study showed higher selectivity to methanol than did Raney copper or copper chromite catalysts under comparable conditions. In addition, copper dispersion was higher, excluding the possibility of the presence of a significant continuous copper surface. Reaction Orders. The reaction orders with respect to methyl formate and hydrogen were determined at 167 "C on catalyst Cu200. The mole fractions (C)in the recycle loop were varied between the following limits: 0.1 < CH,< 0.85 0.1
< CMF< 0.33
From the slope of the resultant log PA vs. log Piplots, a 0.39-order dependency with respect to methyl formate was found. The same reaction order with respect to methyl formate was found at 140 "C. The reaction order with respect to hydrogen was found to be zero over the concentration range studied. The apparent activation energy for reaction 1was then evaluated in the temperature range 400 to 460 K. The experiments were conducted at a fixed ratio of methyl formate:hydrogen of 1:8 and at total inlet flow rates between 100 and 250 cm3/min. The reaction rate constants were expressed in terms of available copper surface area. Excellent linearity was found in the Arrhenius plot at temperatures below 420 K (Figure 1). At higher temperatures the rate constants were much smaller than the values expected from the lower temperature experiments. Calculations based on Knudsen diffusion in the catalyst pores showed that Thiele moduli were well below 1, and typically 0.01. In addition the three catalysts (with different pore structures) showed no difference in reaction
400
Ind. Eng. Chem. Prod. Res. Dev., Vol. 24, No. 3, 1985 -71
I
\
I
I
-
I
-4
A CuOX50
cuzoo
\
I
-14
-I4 1000IT i K-'
I
Figure 1. Arrhenius plot of the rate constant for the hydrogenolysis of methyl formate over catalyst Cu200 without the influence of CO partial pressure considered.
. 4
I
/+--CO
0
introduced
3
0
50
0 97%CO 3,2% CO 5 3 % co
=\
I 3
100 150 200 KO 300 350 CM) 450
T I m e Iminl
Figure 2. Influence of CO on reaction rate as a function of time on stream.
rate constants. As a result, intraparticle mass transfer limitations are not expected to influence the reaction rates measured. Interparticle mass transfer resistances could be ruled out due to the high recycle ratios used that resulted in extremely high gas velocities in the catalyst bed. It was thought that the decline in the reaction rate constants observed at higher temperatures might have arisen from an inhibiting effect by the reaction products methanol and/or carbon monoxide. Therefore the effect of methanol on the reaction rate was investigated in a set of experiments using a methanol/methyl formate mixture in the feed stream. The additional methanol had no influence on the reaction rate. This result was further confirmed by reevaluation of all the kinetic data by multiple linear regression analysis of a power law rate expression incorporating methanol as well as methyl formate and hydrogen partial pressures. The calculations gave the same results for the reaction order with respect to methyl formate and gave further evidence for the zero-order dependence with respect to hydrogen and methanol. The effect of the CO concentration on the reaction rate was investigated in a separate set of experiments. In these runs a 51 mixture of H2:methyl formate was passed over 1g of catalyst Cu200 at a total flow rate of 250 cm3/min for 2 h at 167 OC, and steady-state conditions were established. CO was then added to the reactant gas mixture and the same amount of hydrogen was withdrawn, thus maintaining the residence time constant. The amounts of CO added resulted in carbon monoxide mole fractions in the recycle loop between 0.01 and 0.05. The introduction of CO resulted in an immediate reduction of the reaction rate (Figure 2). The size of the step change in loss of conversion was a function of the CO concentration. A -0.17-order dependence of the reaction rate upon the CO
t
I 21
22
23 1000/T
4
24
6
25
1
26
if' 1
Figure 3. Arrhenius plot of rate constants for the hydrogenolysis of methyl formate with the influence of CO on reaction rate included.
partial pressure was found from these experiments. The experiments with CO in the feed exhibited longer term changes in catalyst activity. The extent of this catalyst deactivation was shown to depend on the CO concentration in the reactor (Figure 2). For CO mole fractions higher than 1-2 % the activity declined continuously after the step change. When a CO concentration of 1% or less was used, the activity also decreased but appeared to reach a constant value after a period of 3 h (Figure 2). Catalyst reactivation following reaction in the presence of added CO at various pressures was studied in a separate set of experiments. In each case the initial activity could be restored by purging the catalyst with pure hydrogen for several hours at reaction temperature. The longer term deactivation in the presence of added CO may shed some light on the deactivation observed by Evans et al. (1983~).As noted earlier, the catalysts used in the previous study were less selective and produced CO levels comparable with those present when CO was added in the current study. Thus CO may be directly or indirectly linked to catalyst deactivation in the hydrogenolysis reaction. A more detailed investigation of the deactivation process in this system using surface chemistry techniques is warranted. Rate Expression. The orders of reaction for methyl formate (0.39) and CO (-0.17) were incorporated in a power law rate expression of the form of eq 6 for the hydrogenolysis of methyl formate over the three catalysts investigated. Figure 3 shows the temperature dependence of the rate constants obtained for all catalysts in the temperature range 397 to 455 K. The activation energy determined from the Arrhenius plot is 117 kJ/mol. The kinetic data for all three catalysts are well described by the single kinetic rate expression rA1= 3.6
X
lo9 exp(-14072/T)pMF0~39pCo-0~17 (8)
Figure 3 shows that the activities of the three catalysts are almost identical when compared on the basis of copper surface area. Since the data were obtained in a recycle reactor, it was quite simple to model the kinetics of the parallel production of carbon monoxide by the decarbonylation of methyl formate (eq 2). A plot of rate of formation of CO vs. methyl formate concentration showed the reaction to be first order in methyl formate. The reaction was also shown to be independent of hydrogen concentration for the range of hydrogen pressures used for the hydrogenolysis reaction. The activation energy for the decarbonylation reaction was determined from data obtained by
40 1
Ind. Eng. Chem. Prod. Res. Dev. 1985, 2 4 , 401-403
using catalyst Cu200. The value of 112 kJ/mol obtained is the same as that for the hydrogenolysis step within experimental error. The rate of the decarbonylation step is described by the following equation: rA2
= 2.2 x
io9 exp(-13471/T)pMF
(9)
The reaction rate constants have units of (mol min-’ m-2cubar4.22)and (mol min-’ m-2cubar-’) for the hydrogenolysis and decarbonylation reactions, respectively. The rate expressions given by (8) and (9) are useful in determining reactor performance for the hydrogenolysis reaction in the two-step route to methanol. They show the importance of producing a highly selective catalyst that minimizes byproduct CO formation. Conclusions The reaction kinetics of the hydrogenolysis of methyl formate have been found to depend on both methyl formate and CO partial pressures whereas hydrogen and methanol partial pressures have no influence on the reaction rate. CO formed by the decarbonylation of methyl formate inhibits the hydrogenolysis reaction and leads to a reversible deactivation of the catalysts at higher concentrations. The hydrogenolysis and decarbonylation of methyl formate have similar activation energies, suggesting that the reaction sequence may involve a common intermediate for both reactions. Copper on silica catalysts prepared by an ion-exchange method have been shown to have high selectivity and activity for the hydrogenolysis of methyl formate. As a result of the low levels of CO produced over these catalysts there is no evidence of the deactivation observed in earlier
studies of this reaction using other copper catalysts. This suggests that these catalysts may have an important use in a two-stage route to methanol comprising carbonylation and hydrogenolysis steps. Acknowledgment Funding of this project under the Australian Research Grants Scheme is gratefully acknowledged. Support was provided under the National Energy Research Development and Demonstration Programme administered by the Department of National Development. Registry No. HC(O)OMe, 107-31-3; M e O H , 67-56-1; CO, 630-08-0; CU, 7440-50-8.
Literature Cited Christiansen, J. A. U.S. Patent 1302011, 1919. Evans, J. W.; Cant, N. W.; Trimm, D. L.; WainwriQht, - M. S. A.m. / . Catal. I983a, 6, 335. Evans, J. w.; Casey, P. S.; Wainwright, M. S.; Trimm, D, L,; Cant, N, w, Appl. Catal. I963b, 7 , 31. Evans, J. W.; Tonner, S. P.; Wainwright, M. S.; Trhnm, D. L.; Cant, N. W., presented at the Eleventh Australian Conference on Chemical Engineering, Brisbane, 1 9 8 3 ~p 509. Evans, J. W.; Wainwright, M. S.; Bridgewater, A. J.; Young, P. J. Appl. Catal.
-~.
1Q83d. .~ 7 . 7 5 .
.
Evans, J. W.; Wainwright, M. S.; Cant, N. W.; Trimm, D. L. J . Catal 1984, 88. 203. Kiier, K. Adv. Catal. 1982, 31, 243. Kobayashi, H.; Takezawa, N.; Minochi, C.; Takahashi, K. Chem. Left. 1980, 1197. Kung, H. H. Cafal. Rev. Sci. Eng. lQ80,22(2), 235. Petrole Informations, 1982, May 13, 34. Sorurn, P. A.; Onsager, 0. T. Roc. I n t . Congr. Catal. 8th 1984, 11-233. Tonner, S. P.; Trimm, D.L.; Wainwright, M. S.; Cant, N. W. Ind. Eng. Chem. prod. Res. D ~ V .1 ~ 8 4 ~ 2 3 , 3 8 4 .
Receiued f o r review November 1, 1984 Accepted April 3, 1985
GENERAL ARTICLES
Evaluation of Manganese Phosphate Coatings Rlchard A. Farrara U S . Army Armament Research and Developent Center, Benet Weapons Laboratory, Watervllet, New York 12189
The corrosion and wear resistance of the basic or normal manganese phosphate coating is compared with that of manganese phosphate that has been modified or converted by a chemical solution named “Endurion”. Supplementary coatings of either oil or solid-film lubrlcant are applied over both types of phosphate.
Introduction The coating commonly used to protect steel from corrosion is the basic, heavy, manganese phosphate and either oil per VV-L-800, [Federal Specification-lubricating oil, general purpose, preservative (water-displacing, low temperature)] or heat-cured solid-film lubricant (SFL) per Mil-L-46010, [Military Specification-lubricant, solid-fib heabcured, corrosion-inhibiting]. This correlates with type M, class 1, of the phosphate specification DOD-P-16232, [Military Specification-phosphate coatings; heavy, manganese, or zinc base (for ferrous metals)]. Oil applied to
basic manganese phosphate is a relatively low cost coating that results in respectable corrosion and wear resistance whereas solid-film lubricant applied to manganese phosphate is a relatively expensive coating that results in good corrosion resistance and excellent wear resistance. The Endurion-modified manganese phosphate (patented process of the LEA Manufacturing Co.), which correlates with type M, class 4,of DOD-P-16232, was selected for comparison with the present manganese phosphate because it has the potential for significant improvement in performance (corrosion and wear resistance) with an associ-
This article not subject to U S . Copyright. PIublished 1985 by the American Chemical Society
