CA1217468A - Removal of metal contaminants from catalysts using buffered oxalic acid - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE

A process for removing metal contaminants from a hydroconversion catalyst, said catalyst containing at least one metal from Groups VIB, VIIB or VIII supported on a refractory inorganic oxide. The process comprises contacting the contaminated catalyst with a buffered oxalic acid solution wherein contaminant is removed without dissolving the support.

Description

7~

2 This invention relates to the removal of metal

3 contaminants from hydroconversion catalysts. More par-

4 ticularly, metal contaminants are selectively removed

5 from catalysts on refractory oxide supports by treatment

6 with buffered oxalic acid solutions without dissolving

7 the support.

8 Hydroconversion catalysts used to treat petro-g leum feedstocks become deactivated due to factors such 10 as coke build-up and contamination by metals typically 11 found in crude oil. It is known to use acids in the 12 process of regenerating hydroconversion catalysts. U~S.
13 Patent 2,380,731 relates to the removal of iron and 14 other metals from catalytic cracking catalysts by treat-15 ing spent catalysts with organic acids and dilute mine-16 ral acids. Aqueous solutions of oxalic acid are pre-17 ferred. U.S. Patent 3,020,239~describes the removal of 18 vanadium from molybdenum containing catalysts using aqu-19 eous glycolic acid solution. Other hydroxy acids and 20 compounds similar to glycolic acid are unsatisfactory 21 because of a concomitant removal of molybdenum. Concen-22 tration control is preferred to avoid leaching of alumi-23 num from support material. U.S. Patent 3,791,989 teaches 24 the removal of vanadium from a hydroprocessing catalyst 25 containing a Group VI or Group VIII metal by contacting 26 the catalyst with an aqueous solution of oxalic acid be-27 fore burning off any coke deposits. In Example 1, 28 fouled catalyst particles were boiled with concentrated 29 oxalic acid solution with a preferential removal of 30 vanadium over nickel or molybdenum. A reverse order is 31 described in British Patent 1,245,358 whereby deactiv-32 ated Group VI or VIII catalyst on a carrier is first 33 subjected to a coke burn~off followed by washing with 34 aqueous oxalic acicl solution having a concentration of 35 0.5 M to saturation. Catalytic metal is not removed 36 provided contact time with oxalic acid is limited.

~).'"

1 ;n U.S. patents 4,089,806 and 4,122,000, hy-2 dro-desulfurization catalysts containing Groups VIB
3 and/or VIII metals on a refractory support are regener-4 ated using a combination of oxalic acid plus nitric acid 5 and/or nitrate salts. ~he use of oxalic acid alone in 6 environments severe enough to remove suostantial amounts 7 of vanadium is stated to result in dissolution of alu-8 mina support. Finally, according to ~.S. Patent

9 3,536,637, iron fouled ion exchange materials are re~u-

10 venated by contacting fouled resin with oxalic acid.

11 While it is known that oxalic acid is useful

12 for removi~g certain metal contaminants from hydrocon-

13 version catalysts supported on inorganic oxides such as

14 alumina, oxalic acid suffers from the disadvantage of

15 also removing or dissolving catalytic metals and support

16 materials.

17 SUMMARY OF THE INVENTION

18 It has been discovered that the above-cited

19 disadvantages may be overcome without any harmful effect

20 on the ability of oxalic acid to remove metal contami-

21 nants if the aqueous oxalic acid solution is buffered.

22 Accordingly, the process of the invention for removing

23 metal contaminants from a hydroconversion catalyst

24 containing at least one metal from Groups VIB, VIIB or

25 VIII supported on a refractory inorganic oxide comprises

26 contacting the contaminated catalyst with a buffered

27 aqueous oxalic acid solution. In another embodiment,

28 metal contaminants are removed from a catalytic cracking

29 catalyst, said cracking catalyst containing at least one

30 of alumina, silica, silica-alumina, zeolites or clays,

31 by a process which comprises contacting the contaminated

32 cracking catalyst with a buffered aqueous oxalic acid

33 solution.

~7~

1 The present process employing buffering per 2 mits the use of oxalic acid at varying concentrations 3 without dissolution of metal oxide supports typically 4 employed in hydroconversion catalysts. Moreover, long 5 contact times for oxalic acid solution with contaminated 6 catalyst are possible without removal of the catalytic-7 ally active metal from support, even at elevated extrac-8 tion temperatures. Buffering oxalic acid also permits 9 its use for regenerating catalytic cracking catalysts 10 without dissolving the catalyst.

11 DETAILED DESCRIPTION OF 'rHE INVENTION

12 ~ydroconversion catalysts of interest are 13 those wherein either the active catalytic component or 14 the support material is susceptible to attack by mineral 15 and/or organic acids In usual commercial use, these 16 catalysts become fouled with metal contaminants and coke 17 formation, leading to catalyst deactivation. Acid 18 regeneration of deactivated catalysts is used to remove 19 metal contaminants. If the support or catalyst is sus-20 ceptible to acid attack, however, there is a substantial 21 problem since loss of support can also lead to loss of 22 active catalytic component. Such losses are frequently 23 unacceptable or contact time between acid and fouled 24 catalyst is limited to such an extent that substantial 25 amounts of metal contaminant cannot be removed.

26 Hydroconversion catalysts of particular inter-27 est include cracking catalysts, reforming catalysts and 28 resid conversion catalysts. Catalytic cracking cata-29 lysts are acidic metal oxides such as alumina, silica, 30 silica-alumina, zeolites, clays, especially acid treated 31 clays, silica-zirconia, silica magnesia, alumina-boria 32 and silica-titania. The combination of the above-cited 33 cracking catalysts as support with a hydrogenation-dehy-

34 drogenation metal, e.g., Co, Ni, W, V, Mo, Pt, Pd or

35 combinations thereof, results in a hydrocracking cata-7~

1 lyst. In general, the combination of a Group VIB metal,2 Group VIIB, Group VIII metal or mixture thereof sup-3 ported on an acidic support as described above are used 4 for a variety of catalytic hydroconversions, such as S resid conversion and reforming. Groups are defined in 6 the Periodic Table on p. 662 of the 9th ed. of the 7 Condensed Chemical Dictionary. Preferred catalytic 8 materials are alumina, silica-alumina or zeolites as 9 supports for Co, Mo, Re or platinum group metals, either 10 alone or in combination.

11 Metal contaminants may arise from either feed-12 stock or metal components used in the fabrication of 13 reactors and transfer lines. Examples of metal contami-14 nants which may be removed by oxalic acid are iron, 15 vanadium, nic'~el, sodium, magnesium~ chromium, copperl 16 strontium, lithium, lead and the like. Such metals are 17 generally undesirable since they may lead to catalyst 18 deactivation by catalyst poisoning, plugging or both.

13 As noted hereinbefore, oxalic acid extractions 20 for removing metal contaminants may also result in acid 21 attack of susceptible supports, active catalytic metals 22 or both. Moreover, some oxalic acid treatments require 23 high temperatures and/or high concentrations of acid 24 thus exacerbating the above-stated problems. It has 25 been discovered that buffering the oxalic acid obviates 26 these problems without interfering with the ability of 27 oxalic acid to extract metal contaminants.

28 Buffering agents according to the invention 29 are those which are capable of buffering aqueous oxalic 30 acid solutions within a pH range of from 2 to 10, pre-31 ferably 4.5 to 8.5. Preferred buffering agents are 32 those which do not contain a Group IA or ~roup IIA ca-33 tion in an amount sufficient to cause catalyst poisoning~

~2~7'~

1 Examples of preferred buffers are ammonium 2 salts of weak acids, e.g., ammonium acetate, ammonium 3 citrate, ammonium formate, ammonium lactate, ammonium 4 valerate, ammonium proprionate, ammonium urate, etc., or mixtures thereof. Mixtures of different buffers are 6 known to extend the range of pH which can be controlled 7 compared to the separate buffers. The use of the 8 ammonium salts of weak acids precludes the possible g adverse effect of adsorption of the cations of alkali and alkaline buffers. Such cations are known to be 11 severe poisons for many hydroconversion catalysts.

12 The use of metal chelating buffers may also 13 be used to aid in the selective removal of metal con-14 taminants from hvdroconversion catalysts. Oxalic acid itself is known to chelate and to remove niobium and 16 zirconium from ion exchange columns ("The Chemistry of 17 Lanthanides", T. Moeller, Reinhold Publishing Corpora-18 tion, New York, 1963, p81-853. Buffers which have dif-19 fering metal-binding abilities are: NH2COCH2NHCH2CH2S03H, NH2COCH2N(CH2COOH)2, bicine, glycylglycine, O ~ CH2CH2S03H, 21 tricine. ~"Buffers for pH and M~tal IOD Control," D. D.
22 Perrin and B. Dempsey, John Wiley and Sons, ~ew York, 23 plO8). Glycine derivatives are known to be strong com-24 plexing agents toward iron and other metals. Hexanoic and Di(o-hydroxyphenyl) acetic acid are also known to be 26 strong complexing agents for iron and other metals, and 27 may be added to buffer compositions ("Photometric and 28 Fluorometric Methods of ~nalysis. Metals Part 1," F. D.
29 Snell, John Wiley and Sons, New York, 1978, p763).
Enhancement in the rate and extent of metal contaminant 31 removal from a hydroconversion catalyst by utilizing a 32 strong complexing agent in the bufer solution, in addi-33 tion to oxalate ions, can lead to an improved metal 34 contaminant extraction scheme.

Another class of buffers for metal contaminant

36 removal from hydroconversion catalysts involve buffers 1 useful in aqueous/non-aqueous solvent systems. For 2 example, in a 90% MeOH/10% H2O solution a 0.01 molar 3 oxalic acid/ 0.01 molar ammonium oxalate buffers the pH*
4 to 4.23. In a 60% MeOH/40% H2O solution this same buffer gives a pH* of 2.58. (See "Buffers for pH and 6 Metal Ion Control," D. D. Perrin and B. Dempsey, John 7 Wiley and Sons, New York, lg74, p84-88.) Aqueous/non-8 aqueous systems are useful for metal contaminant removal 9 from hydroconversion catalysts as the alumina solubili-zation is greatly inhibited compared to aqueous systems.

11 In certain cases, buffers may contain small 12 concentrations of Group IA component. In this case, 13 such buffer systems may be useful to remove metal con-14 taminants from hydroconversion catalysts. ~xamples of such aqueous buffers are phenylacetic acid and sodium 16 phenylacetate, acetic acid and sodium acetate, succinic 17 acid and sodium hydroxide, potassium hydrogen phthalate 18 and sodium hydroxide, sodium hydrogen maleate and sodium 19 hydroxide, potassium dihydrophosphate and sodium hydro-xide, sodium pyrophosphate and hydrogen chloride, bicine 21 and sodium hydroxide. All of the above buffers have a 22 pH range which is defined by the buffer component 23 concentrations.

24 Other aqueous buffers useful in conjunction with oxalic acid in removing metal contaminants from 26 hydroconversion catalysts include such aqueous buffers 27 as glycine and hydrogen chloride, imidazole and hydrogen 28 chloride, 2,4,6-trimethylpyridine and hydrogen chloride, 29 N-ethyl-morpholine and hydrogen chloride, tris(hydroxy-methyl)-aminomethane, 2-amino-2-methylpropane-1,3-diol 31 and hydrogen chloride, diethanolamine and hydrogen 32 chloride, and ammonia and ammonium chloride.

33 In some instances, a wide-range bufEer (pH
34 2.6-8~0) such as citric acid and Na2HPO4 may also be useful in removing metal contaminants from hydroconver-1 sion catalysts. Other well-known wide-range buffers 2 are: piperazine dihydrochloride, glycylglycine and 3 sodium hydroxide; citric acid, K2HPO4, diethylbarbituric 4 acid and sodium hydroxide, boric acid, citric acid and tri-sodium orthophosphate.

6 The amount of buffer required is dependent on 7 the concentration of oxalic acid employed. Oxalic acid 8 concentrations may range from 0.0001M to 5.0M, and con-g centration ranges of from 0.01 to 2.0M are preferred.
The amount of buffer required is that sufficient to 11 maintain the pH of oxalic acid within the desired range.
12 Increasing the concentration of oxalic acid will gene-13 rally require greater amounts of any given buffer~ de-14 pending on the buffering capacity.

Temperature ranges are from about 0C to about 16 100C, preferably 20 to 75C. Higher temperatures may ]7 lead to excessive acid attack on the support and/or 18 catalytic metal without any particular benefit with 19 respect to removal of metal contaminants.

Unlike conventional oxalic acid, buffered 21 oxalic acid can be contacted with deactivated catalysts 22 for prolonged periods without dissolution of support 23 materials. The manner of contacting catalyst and 2~ buffered oxalic acid solution i5 not critical. Deac-tivated catalyst may be soaked in buffered oxal;c acid solution. Samples of catalyst or buffered oxalic acid 27 solution can then be removed at fixed intervals to 28 monitor extent of metal decontamination. Alternatively,29- catalyst may be continuously extracted, for example, in a countercurrent extractor with buffered oxalic acid 31 recycle.

32 A preferred embodiment for removal of metal 33 contaminants from hydroconversion catalysts relates to 34 the regeneration of reforming catalysts which typically :

~2~

1 contain Pt eithe~ alone or in combination with another 2 metal, preferably Re or a noble metal, on an acidic 3 support such as alumina or silica-alumina. Reforming ~ catalysts gradually become contaminated with metals such as Fe, Pb, Cu, Ca, Na and the like with Fe being parti-6 cularly troublesome.

7 When 0.5 M oxalic acid is slurried with Fe 8 contaminated Pt, Pt/Ir or Pt/Re catalysts on ~A1203 and g heated, substantial amounts of support and noble metal are removed along with the desired Fe contaminant. On 11 the other hand, an ammonium acetate buffered oxalic acid 12 extraction of the ~ame catalyst systems results in a 13 selective removal of Fe with only very small losses of 14 catalyst which are attributed to mechanical losses rather than chemical attack on the ~-alumina support.
16 For Fe removal using buffered oxalic acid solutions, 17 lower temperatures of from 20 to 75C are favored.

L8 Following extraction, a Pt/Re catalyst was 19 calcined in the presence of 2 at 500C for four hours.
H2 and CO chemisorption values of regenerated catalyst 21 were nearly the same as those of fresh catalyst indicat-22 ing no poisoning or surface modification by buffered 23 oxalic acid treatment.

24 An Fe extracted Pt/Ir catalyst was compared against the non-extracted catalyst containing 0.35 wt.%
26 Fe and a fresh Pt/Ir catalyst for naptha reforming 27 activity. ~fter 400 hours, the extracted catalyst 28 showed only a 2~% lower reforming activity relative to 29 the fresh catalyst compared to a 70% decline in reform-ing activity for the catalyst which had not been ex-31 tracted with buffered oxalic acid solution.

32 The process of the invention is further illus-33 trated by the following examples.

~L2~L~7L~

Catal~sts 2 Catalysts were supported on ~-~1203 supports 3 having ~ET surface areas of from 150-190 m2/g.

4 Pt/Re/A1203. The platinum-rhenium bimetallic catalyst was a commercial sample with an initial compo-6 sition of 0.3% Pt, 0.3% Re and 0.9% Cl (all percentages 7 are weight percent unless otherwise indicated).

8 Pt/Ir/A1203. Platinum-iridium bimetallic g catalysts employed were commercial samples. The fresh catalyst contains 0.3% Pt, 0.3% Ir and 0.7% Cl. A
11 number of used Pt/Ir/A120J samples were also ob-12 tained from commercial refineries.

13 Pt/A1203. The monometallic platinum catalyst14 was a commercial sample with a composition of 0.3% Pt and 0.7% Cl.

16 Iron Doped Catalysts. Known quantities of Fe 17 were added to fresh Pt/Re/A1203 and Pt/Ir/A1203 cata-:L8 lysts by an incipient wetness procedur~ using standard-19 ized, aqueous iron nitrate solutions. After drying under air at 120C for 16 hours, the impregnates were ~1 calcined at 270C for four hours under 20% 02/~e (SOOcc/
22 min.) to insure decomposition of the nitrate salt.

23 Example 1 24 This example demonstrates the harmful effects of Fe on the chemisorption properties of noble metal 26 bimetallic catalysts. Hydrogen and carbon monoxide 27 chemisorption studies were performed with a conventional 28 glass vacuum system as described by ~infelt and Yates in 29 J. Catal., 8, 82 (1967).

~7~

1 Hydrogen and carbon monoxide uptakes were 2 determined at 25 + 2C on the reduced and evacuated sam-3 ples. Typically, 30 minutes were allowed for each up-4 take point. H/M ratios were calculated by assuming that hydrogen uptake at zero pressure corresponds to satura-6 tion covera~e of the metal. Hydrogen uptake at zero 7 pressure was determined by extrapolation of the high 8 pressure linear portion of the isotherm as described by g Benson and Boudart and Wilson and Hall (J. Catal., 4, 704 (1965) and 17, 190 (1970). CO/M ratios were calcu-11 lated by determining the carbon monoxide uptakes on the 12 reduced and evacuated samples and assuming that this 13 quantity represented the sum of carbon monoxide weakly 14 bound to the A1203 support and strongly bound to the metal surface. The sample was then evacuated (10-5 Torr) 16 for 10 minutes at room temperature and a second carbon 17 monoxide isotherm measured. Since the second isotherm 1~ measures only the carbon monoxide weakly adsorbed on the 19 support, subtraction of the two isotherms gives the quantity of carbon monoxide strongly associated with the 21 metal components. The amount of strongly bound carbon 22 monoxide at 100 Torr was chosen as saturation coverage 23 of the metal.

24 The hydrogen chemisorption properties of fresh and iron doped Pt/Re and Pt/Ir catalysts are summarized 26 in Table 1.

~2~L7~

2Hydrogen Chemisorption Properties of Iron Doped 3Pt/Re and Pt/Ir on Alumina Catalysts____ 4 Catalyst% Fe(a) H/M(b) D/Do(C) 0.3% Pt/0.3% Re/A1203 0 0.4 100 6 0.10 0.3 75 7 0.36 0.2 50 8 0-3% Pt/0.3% Ir/A12O3 0 1.7 100 9 0.05 1.4 82 0.10 1.3 76 11 0.36 1.0 59 12 0.80 0.8 47 13 a) Iron added as the nitrate salt. Following iron 14 impreqnation the catalysts were calcined at 270C
under 20% O2/He (500 cc/min) for 4.0 hr to ensure 16 decomposition of the nitrate salt.
17 b) Atoms of hydrogen chemisorbed per metal atom in the 18 catalysts at room temperature. Prior to chemisorp-19 tion measurements, the catalysts were reduced at 500C under H2 (500 cc/min) for 2.0 hr.
21 c) Normalized dispersions 22 In the last column of Table 1, the hydrogen 23 uptakes of the iron doped catalysts have been normalized 24 with respect to the uptakes of the fresh Pt/Re and Pt/Ir catalysts. Systematic decreases in hydrogen chemisorp-26 tion capacity for both bimetallic catalysts were ob-27 served with increasing iron concentrations. At an iron 28 concentration of 0.36%, the hydrogen uptakes of the 29 bimetallic catalysts are 40-50% lower than those exhi-bited by fresh samples.

~7ql~

1 The Fe/noble metal mole ratios for catalysts containing 2 0.1 to 0.8% Fe range from 0.5 to 4. A 0.2% iron concen-3 tration would thus be high enough to interact with all 4 the noble metal components present in Pt/Re and Pt/Ir reforming catalysts. The marked decreases in H/M values 6 caused by low iron concentrations suggest in fact that 7 iron is either alloying with or physically blocking a g sizeable fraction of the noble metals. A decreased g chemisorption capacity could potentially lower reforMing activity since H-H and C-H bond activation processes are 11 extremely important in catalytic reforming.

12 Example 2 13 The effects of iron on the reforming activi-14 ties of bimetallic catalysts is shown in this example.
Naphtha reforming reactions were carried out in a 25cc, 16 stainless steel, fixed-bed, isothermal hydrotreating L7 unit operated in a single pass mode.

18 Reforming experiments were performed at 487-19 489C under 200 psig total pressur~. The weight hour space velocity was 2.1 WHW and hydrogen was supplied at 21 a rate of 6000 SC~ H2/BBL. A commercial naphtha feed-~2 stock was utilized and the sulfur concentration adjusted 23 to 0.5 ppm by the addition of thiophene. Following hy-~ drogen reduction at 500C, the catalysts were s~lfided using a dilute H2S/H2 mixture. Reformate was analyzed 26 for research octane number tRO~). Octane numbers were 27 used to define relative catalyst activities (RCA).

28 The naphtha reforming activities of fresh and 29 iron doped Pt/Re and Pt/Ir catalysts are compared in Table 2.

~L7'~

C~

~ ~ ~ ~ ~ n ~ ~ o ~
o tn _ ... .. ... ..
Q~ z ~ ~ u~ o ~,1_I 1`~
u~ O ~ cn tn ~ ~ o ~ oG~
C ~C~
o~
H
o o ~
o a~ co ao ~ ~ ~ u~ 9 Ln Q
U~ ~ U~ D CO O ~ CO
~~; ~ ~ ~ ~~1 ~ ~ ~ ~ ~1 ~E~
.,1 ~ C , .,1 _I ~ .,~
~ .
m ~: " cl~ . . ~ ~
E~ ~: o o C
C
.~, ~ ~ ~ ~
E D. ~ v o ~ ~ o O O O O
~;
~ -- ~ ~ ~
~;
~ ~ --JJ ~ ~ ~ ~ H H 111 U~ _I
ul J~
~1 ~1 0~0 o\O o~O o~O
U
æ ~
~ O O O O
C~ ~ I~J ~ ~ ~ O ~
OP 0\o 0~o 0\o ~a ~ u O O O O ------L74~

1 Relative catalyst activities clearly show that 2 0.36% iron has a pronounced deleterious effect on both 3 of the bimetallic catalysts. Similar coke makes were 4 found for all the catalysts. The 55-65% lower activi-ties exhibited by the iron containing catalysts indic-6 ates that the noble metals are deactivated by interac-7 tion with iron. Since the Fe/noble metal mole ratio is 8 near 2 in these catalysts, there is enough iron avail-g able to modify all the Pt, Ir and Re. The decreased reforming activities are in agreement with the reduced 11 hydrogen chemisorption capacities found for the iron 12 doped catalysts.

13 Example 3 14 The solubilities of ~ -A1203 in H2ol oxalic acid and ammonium acetate buffered oxalic acid solutions 16 are compared in Table 3.

~7~

_ .
o\o . ~ o _ O cn o . ~
ra m o o) ~ U~
o J~
o ~ c~ 0 o . ~ o ~ ~
~) I'S U') ~ ~: E ~ O
C~ .
O ~ ~ X
~ O O) O :~
U- S
C V
C ~: o o C
~ Ol O O O
X~(C~l O O L) U~ ~
O ~ ~ ~ ~ ~ ~ ~D 0 ~5 ~ ~ U~ O
U~ O ~\ S
~s o U~
o u~ ro ~ ~D
In~rl ~ O
V o 4 ~1 O ~ ~
v a) Ei S O
V
~ . ~rl O ~rl E:l~' V ~'C 0 _I ~ 3 t~
1 l~ ~ ~ ~ O
m ~ o . . ~x ~~o O ~ .d' '1' E; O s ~ rJ
E~~ X ~ ~U~
o 3: o :~:C
¢ ~ u~
I ~ .. ,~
Q) C
~ o .,, 4~ ~ ~ ~ ~ O ~5 ~ ~ ~ .
O ~1 ~ . ~ l 3 ~J
~ ~_1 O O N X ~ ~
~ m o ~ U~ ~I ~ ~ C
.,, _I c a~
U O ~1 o ~O~ _ . .
:: ~ ~O ~-- a; ~
O V N~rl ~ O tll U~ 1¢ ~1 C) I~S
~
o a~
V ~ U :I:
,1 a~ Q) O
. O X ~ V~
I~ ~Ul O O ~) (I~
_ ~ a u c o Ir) a) G) U~
a) v S~r~
E~ 1~ o 1 a) oO O _ U O O O ~ ~
E~ E~,.~ ,.1 _ _ ~Z~7~

1 After refluxlng in H2O for 30 min., the pH of 2 an ~ -A12O3 slurry reaches a self-buffering level of 3 about 5Ø The acidic nature of A12O3-H2O slurries is 4 well-known in the art. A slurry of A12O3 in oxalic acid exhibits a pH of around 2 at 100C~ Thus, the acidity 6 f an A12O3-oxalic acid slurry is approximately three-7 orders of magnitude greater than that of an A12O3-H2O
8 slurry. An A12O3-ammonium acetate buffered oxalic acid g slurry, in contrast to the oxalic acid slurry, was ob-served to maintain a relatively constant pH value of 11 around 4.4. Thus, the pH of the buffered slurry is 12 similar in value to that of an aqueous slurry of A12O3.
13 The effects of pH are reflected in the quantities of 14 A12O3 which can be recovered unchanged from the slurries Essentially complete recovery of A12O3 was achieved from 16 the H2O and buffered oxalic acid slurries. Oxalic acid 17 slurries, however, readily dissolve substantial quanti-18 ties of A12O3. After refluxing for 30 min. nearly half 19 of the starting quantity of A12O3 was transformed into a soluble aluminum species.

21 Example 4 -22 This example illustrates the non-selective 23 extraction of iron from noble metal catalysts. The 24 results of representative studies using oxalic acid solutions for e~tracting iron from Pt, Pt/Ir and Pt/Re 26 catalysts are presented in Table 4.

~L7~

coo ~ a7 O ~1 0 0 ~ ~10 ~a ~ .. .. ..
~_ o oo o o ~o ~, u~ ~11 1 1 11 1 o D:;lI I I II I -o ~, ~ ~ ~
a) u~ O
U~ ~ ~ ~ ~ ,` Ln C, O ~.:I I~ ~ N N I O U~
O O .~ o 1_1 ~ ~ O O O O W U~ O
~ ~ aJ o a~
O ~ ~V
~_~ ~ 3 u~
o x-- c~In o u~ C
~ U~ ~ ~JN ~1 -1 ~I N --1 N O
tO

~-r1 " X
~ V ~ ~ ~R
~ /3 O O
O ::~
t) _ ~) _I O
~I-ri ra U~
0 ~ U~ ~ O ~ ~D
E x Q) :~ ~P o ~ rl ,1 O O ~ _I .O N~ CO O O1`
~, al (tl ~cl~ ~ I~ ~c~ I` co ~ ~
W U~ ~ ~ ~ ~ 0 ~ O
C O C~ C~ ~ W rl o o Q) E~ 0 3 o~
H cr O E ~ O
1 1~1 11~ N
m41 o ~l ~~ ~ - cw c E~C ~ Co oo o o oo rl r~ ~
C ~1 _ ,1~ ~D~I N U~ 1 0 o tO ~ E~ ~ 5,C
~1 ~ C aJ J~
~) .IJ ~1 0 o ~ ~ a c 3 ~1 ~
(~ S ~ 3 J C~ O O O OO O O t~ U~
X -1 ~1 OIn O O OLl~) O 11~ ~ ~ ~ C: N
G~ l E~ E~ ~1 ~ I,q Iq aJ O
~ . ' ~ ~S
O
:' V ~ 0 ~ H
.,, Lr~ O 1` ~ ~ ~ ~a _, a~ I ~ ct~o t~ ~ ~_ ~ .) U 1 O~ ~ Vl rJO O N O l~
_I a) :>~ w ~ o tD ~ _I _ O ' O N U~
U~ ~I tQ cn ~ ~ ~1 rD I I I N E~
c _l ta 115 ~;I I I ~ O
o A t ~ o ~ o o , Z o ~ o r) _~
Z ~ ` V
c ~ r~ Ln-~
~1 oP ~1 l I N N I I ~ o 1-11 1 I O a~ a) tQ ~1 o O ~ X S-~
~ 3 U~--C~
~J¦ N N N N _ ~_ ~:4 1 ' ' ' ' ~ Q t) O O O O `' -- ~

N ~ ~ U) ~D~ c0 ~ O~I N ~ d' ~ ~D 1` 00 ' 1 After 60 min. at 50C, approximately 10% of 2 the A12O3 support is solubilized by a 0.5M oxalic acid 3 solution. Extractions carried out at 100C dissolve 30-4 40% of the A12O3 support in 60 mins. These substantial support losses are too high for oxalic acid to be seri~
6 ously considered as an extraction agent for metal conta-7 minants. During extractions carried out at higher tem-8 peratures (ca. 100C), a sizeable fraction of the noble g metal components are co-extracted alon~ with the iron.
The loss of Re is particularly noticeable and cannot be 11 controlled by lowering the extraction temperature or 12 decreasing the concentration of the oxalic acid solu-13 tion.

14 Example 5 In contrast to Example 4, selective extraction 16 of iron from noble metal catalysts using buffered oxalic 17 acid is shown herein. The results of studies using 18 ammonium acetate buffered oxalic acid solutions for ex 19 tracting iron from Pt, Pt/Ir and Pt/Re catalysts are summarized in Table 5. Prior to extraction with buf-21 fered oxalic acid solutions, the iron contaminated 22 catalysts were typically reduced under H2, ~2~

o CO
~1 o ~~1 ~ o~ u~ ~ o ~ - -U~ o o ~ ,1 o ~J ~ o JJ_ U~
cn ~1 1 1 1 1 U~ ~; I I I I I I I I
o CO O O 0~ C
I I ~ ~ r~
a) ~
\o O O O O O X ~D O
C~ o o ~ ~ U~
~n ~ ~
3 u~ o o-~ x-- o~ o o o es ~ co o o o &
æ~ ~ ~ ~ ~ ~ .
,c ~ .. . .... . 4~ e , O O O O O O O O O
a) ~ o ~o U~
~ c E
o ~ o ~X _ ,1 s o c~o ~ ~ ~ o u~ a ~ _ .IJ
U~
a) ~ dP 0-~
~ ~ ~ ~ . 0~ 0 CO ~ ~Q ~ W
c ~ ~ ~ ~ ~ ~ ~ ~ a~ o ~ ~ U~ ~rl .,~ ~ :~ ~ 3 ~ ~n ~ ~ s:
~nE ~ o ~ _ o ~ ~ o ~q a) a ~~: ~ " ,~
m E~ ,_ _ ~ 0 o o E~ O ~ ~ ~ o ,/
~ ~ _ _, 4J aJ t~ C o o oo o o oo E ~--I E
t) _ ,~ ~ ~ C
c ,s: ~ E o ~
o ~ Lr ~ ~ ~_ .
E _ _ c ~ a~
.,~
C
o o a) c~ o o o o o o o o ~ ~ 3 ~J ~
E~ O u~ o u~ ~ E ta c I ~o O-r~U~ a) o -O ~ rC ~~ C ~ O
aE~.~ a)l ~ -~ o ." ~ ~ '`~
u~ ~ . .. . u~
:~ o o~ O ~ c u~
<u ~ m ~q~
.,~ ~ ~_ o~ v u~ ~ 3 t~ _~ _ a r~ ~ o~ I I I c~ o (U r-l IJ ~1 O tll a) U! r~
~1 ~ 1~ ~a 4.
~ ~ u ~ o ~ ~ o a~
U~ ~ ~ ~ U ,1 aJ--I
C~ ~ ~1 1 ~~
,1 o\~ oo ~ ~ ~ ~ o o-.~ 3 U X 5~ o s~ a) E~
~n-- a~
_~
. . ~ ra . U~
o oo o _ _ __ 1 AS compared to oxalic acid slurries, buEfered 2 oxalic acid solutions do not solubilize the A1203 sup~
3 port. The apparent catalyst losses of 2-4% are due pri-4 marily to mechanical losses upon filtering the slurries rather than from solubilization of A12O3 in the buffered 6 solution. Iron is selectively extracted and no apparent 7 removal of noble metal occurs even at extraction tempera-8 tures of 100C. The extraction data further indicates g that iron removal is favored at lower treatment tempera-tures. This behavior is consistent with the fact that 11 iron oxalate complexes are more stable at lower tempera-12 tures.

13 The effect of catalyst particle size on ex-14 traction efficiency is shown in the case of a 2% iron doped Pt/Ir catalyst. Smaller catalyst particle sizes 16 would be expected to expose more surface to the extrac-17 tion media. Thus, under a given set of extraction 18 conditions, the removal of iron and other extractable 19 contaminants should become more favorable with decreas-ing particle sizes. The enhanced removal of iron from 21 the powdered catalyst is in agreement with these argu-22 ments.

23 The extraction data reported in Table 5 were 2~ obtained from crude slurry experiments and were not optimized. It is reasonable to assume, however, that 26 essentailly all the iron present on a particular cata-27 lyst could be removed by either recycling the catalyst 28 to fresh extraction media or by devising an extraction 29 procedure whereby the catalyst is continuously contacted with a buffered oxalic acid solution. The use of buf-31 fered oxalic acid solutions can also be applied for the 32 removal of iron and other extractable metals such as Na, 33 V, Ni, Cu and the like from a wide range of hydrocarbon 34 conversion catalysts. While oxalic acid removal is preferred for metal contaminant removal, it may be 36 possible to substitute other organic acids in the

37 process of the invention.

1 Example 6 ~ It is known that iron can act as a poison for 3 noble metal catalysts. This example demonstrates the 4 beneficial effect of iron removal on the chemisorption behavior of a Pt/Re catalyst. The results are shown in 6 Table 6.

o~ a 6 ~ 15 e E ~ ~ x o ~) ~ S
E x ~ ~) O 1~ o o ~ ~) o o ~
O ~1 0 ~ ~1 0~ ~ r~ O ~
^ e ,~ u- 0 In ~n ~ ~ 3 ~
.,, ~ ~ m-~ x ~~ ~ ,, x ~ ,~ c :c E ~ --c o ~ E ~s c o E ~n o E ~ u~ o O~ ~ ~' E ~ c 0 ~4 ~ E ~ C a~ ~
tn c~ E ~ o~ o u~ 6 a~ o E3 u ,CU~ ~ ~ a1 ~C ~~ a ~ S ~ ~~nC Q~~ C ~ ~
~ ~ O ~ ~ ~ O
C~~4~ v C~ 3 R u~ ~4 H U~ t~ 3 R m R O
~ ` ov a) m ~ .,, ~ o _ ~ o v ~Q u~ 0 u~
o ~: ~a~ ~ ~ S
.,, v O _~o o o o o 0-,~
x ~ ~ t~ 0 S E
~5 v~
O ,, c _I x o O--I ~ O J~
H ~ 0 O ~-1 O
O L
~
~D O a) ~1 o ~
U~_I ~ ~ . . . . . . ,~
E~ 0 R R ~ ~ ~ o o o o ~ ,~ o ~ ~ .
m vz o v ~ ~
~ U U~
E~ a ~Q~ ~ O
o ~
~J ~J ~ ~ O
o a) 0o ~ In O ~` CO ~ Ql O
c c~ a) I ~ ~ I ~ o ~ E o o ~ 14 I . . I
,, u~ o o o o a~
O
~ c E o o-,~ ~ , o ~ ~ _1 3~ cn ~ ~ a .,~ ~ ~ 011 1 1 ~ ~ N C V
~i -I ~ ~1 1 ~ I 115 ~J O O O
~ ~ ~
O U~
. ~D CO )-I :~ Ll ~I r~ ~ ~ ~ ~ 0 3 H I . . . I I I >1 tl5 3 o o o ~ I I I S J~
n ~ ~
~Q o ~n ~ c~ o ~ o 0 _~ ~ ~~ ~ ~ ~ ~ ~ S
~ P~
~ oo o o o o o ~
la v G
C) a ~
_ ~7~

1 Addition of 0.37% iron to a fresh Pt/Re cata-2 lyst reduced the H2 and CO uptakes by about 50%. Ex-3 traction of the iron doped catalyst with an ammonium 4 acetate buffered oxalic acid solution at 50C lowered the iron concentration from 0.37 to 0.08%. The ex-6 tracted catalyst was calcined at 500C under 20% 02/He 7 for 4 hours to insure removal of any last traces of 8 extraction agent from the catalyst surface. Following 9 the extraction-calcination steps, the chemisorption uptakes were increased to values near those exhibited by 11 the fresh catalyst. Restoration of the chemisorption 12 values of the Pt/Re catalyst upon iron removal indicates 13 that the buffered oxalic acid extraction agent neither 14 poisons nor modifies the metal surface.

The redispersed Pt/Ir catalyst exhibited hy-16 drogen and carbon monoxide uptakes lower than those 17 given by a fresh catalyst. X-ray diffraction measure-L8 ments on the redispersed catalyst did not exhibit any Pt 19 or Ir diffraction patterns, thus it is reasonable to assume that the catalyst is well-dispersed. The pres-Zl ence of 0.35% iron on the redispersed catalyst may 22 therefore be responsible for the anomolously low hydro-23 gen and carbon monoxide uptakes~ Extraction of the iron 24 from the catalyst with an ammonium acetate buffered oxalic acid solution lowered the iron concentration ~rom 26 0.35 to 0.10%. Following the standard extraction p~oce-27 dure, the catalyst was calcined at 270C under 20% O2/He 28 for 4 hours. The iron extracted and calcined Pt/Ir 29 catalyst exhibited hydrogen and carbon monoxide uptakes which were somewhat below those given by the redispersed 31 catalyst. Since the iron extracted catalyst was sub-32 jected to only a 270C calcination step, the low chemi-33 sorption values may result from residual quantites of 34 the extraction agents adhering to the metal surface.
The low chemisorption values displayed by the extracted 36 catalyst did not, as will be described subsequently, 37 affect the reforming activity of this catalyst. This 7~

1 finding suggests that residual extraction agents can be 2 removed from the catalyst at the higher reforming tem-3 peratures.

4 Example 7 The naphtha reforming activities of a redis-6 persed Pt/Ir catalyst ~containing 0.35% Fe) is compared 7 in this example with a fresh Pt/Ir catalyst and one with 8 the Fe extracted. The results are shown in Table 7.
g The reforming reaction conditions are described in Example 2.

2Naphtha Reforming Activites of Iron Extracted, 3Alumina Supported Pt/Ir Catalysts 4Hrs on Catalyst(a) Feed RON(b) RCA(C) 6 (A), fresh catalyst 72 103.1 176 7 143 102.2 143 8 215 101.5 124 9 335 101.8 134 ~55 101.6 126 11 (B), Redispersed catalyst 12 containing 64 99.9 91 13 0.35% Fe 136 97.2 60 14 208 95.0 48 328 93.3 40 16 400 9~.6 39 17 (C), iron extracted 65 104.7 258 18 from catalyst (B) 137 103.1 176 19 209 102.5 155 305 101.2 117 21 401 100.3 97 22 ~49 100.3 97 23 (a) See Table 6 for complete catalyst specifications 24 (b) RON = research octane number (c) RCA = relative catalyst activity 26 After 400 hr on feed, the redispersed catalyst 27 displayed a reforming activity only about 30~ of that of 28 a fresh catalyst. The reforming activity debit shown by 29 the redispersed catalyst suggests that iron contamina-tion may be an important factor in the deactivation~
31 Following the iron extraction, the reforming activity of ~l 2 ~ 7 ~
~ 26 -1 the extracted catalyst, after 400 hr. on feed, is about 2 80% of that of a fresh Pt/Ir catalyst. Thus, removal of 3 iron from the redispersed sample yields a markedly more 4 active catalyst.

The detrimental effect of a 0.37% iron on the 6 reforming activity of a Pt/Re catalyst is shown in Table 7 8.

~2~6~

2Naph.ha Reforming Activites of Iron Extraced, 3Alumina Supported Pt/Re Catalysts 4Hrs on Catalyst (a) Feed RON(b) RcA(c) 6 (D), fresh catalyst65102.1 142 7 137100.3 93 8 20898.5 72 9 32997.~ 61 40196.9 57 11 (E), iron doped 6399.2 80 12 catalyst containing135 96.0 53 13 0~37~ Fe 20794.1 45 1~ 27891.7 36 32690.0 32 16 35089.5 30 17 (F), iron extracted64102.5 155 18 from catalyst (E) 136 101.3 119 19 20899.9 91 32798.5 72 21 39998.3 66 22 42397.4 61 23 ~a) See Table 6 for complete catalyst specifications 24 ~b) RON = research octane number (c~ RCA = relative catalyst activity 26The reforming activity of the iron doped cata-27lyst, after 350 hours on feed, is only 50% of that of 28 a fresh catalyst. The same iron doped catalyst demon-29 strated 50% lower hydrogen and carbon monoxide uptakes than a fresh catalyst (see Table 6). Thus, excellent 31 agreement between catalytic activity and chemisorption 1 measurements exists for Pt/Re catalysts contaminated 2 with iron. Extraction of 80% of the iron from the 3 iron doped Pt/Re catalyst yields a catalyst which 4 exhibits hydrogen and carbon monoxide uptakes and reforming activity near those shown by a fresh catalyst.
6 The agreement between chemisorption values and catalytic 7 reforming activites indicate that iron poisons bimetal-8 lic catalysts by interacting with the metal components.
9 The fraction of noble metals rendered inactive by iron increases with increasing iron concentration.

11 Example 8 12 This example shows that mineral acids are not 13 equivalent to oxalic acid for iron removal from cata-14 lysts. A catalyst containing 0.28% Pt, 0.27% Ir and 2.0% Fe on an A1203 support was treat~d with hydrochloric 16 acid and hydrochloric acid buffered with ammonium 17 acetate with the results shown in Table 9.

19 Recovered Catalyst 20 Treatment(a) Recovered(b) (Wt.% metals) 21 Solution pH Catalyst (Wt.%) Pt Ir Fe 22 50 ml. of 0.5 MHCl 1.0 83 0.25 0O26 1.6 23 50 ml. of 0.25M 3.0 93 0.27 0.27 1.9 24 ~Cl/0.25 ammonium acetate 26 (a) 3.0 g. of catalyst (1/16" extrudates) were heated at 27 100C for 60 minO

28 (b) Recovered catalysts were rinsed with five 100 ml.
29 aliquots distilled water and dried at 120C for 16 hours, 31 These results demonstrate that buffered HCl 32 solutions cause large support losses without significant 33 removal of iron.

Claims (19)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE PROPERTY OR PRIVILEGE
IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A process for removing metal contaminants from a hydroconversion catalyst containing at least one metal from Groups VIB, VIIB or VIII supported on a refractory inorganic oxide which is susceptible to acid attack which process comprises contacting the contaminated catalyst at a temperature from about 0°C to 100°C with a buffered oxalic acid solution having a pH from about 3 to about 10, wherein the oxalic acid concentration is from about 0.0001M to 5M.

2. The process of claim 1 wherein the oxalic acid solution is buffered to a pH of about 3 to about 8.5

3. The process of claim 2 wherein the oxalic acid concentration is from about 0.001M to 2M.

4. The process of claim 1 wherein the oxalic acid solution is buffered with an ammonium salt selected from the group consisting of ammonium acetate, ammonium citrate ammonium formate, ammonium lactate, ammonium valerate, ammonium proprionate, ammonium urate, and mixtures thereof.

5. The process of claim 3 wherein the oxalic acid solution is buffered with an ammonium salt selected from the group consisting of ammonium acetate, ammonium citrate, ammonium formate, ammonium lactate, ammonium valerate, ammonium proprionate, ammonium urate, and mixtures thereof.

6. The process of claim 5 wherein the temperature of contacting is from about 20°C to about 30°c.

7. The process of claim 1 wherein the catalyst contains at least one metal selected from Co, Mo, Re and the Pt group metals and the support is alumina, silica alumina, or a zeolite.

8. The process of claim 6 wherein the catalyst contains at least one metal selected from Co, Mo, Re and the Pt group metals and the support is alumina, silica alumina, or a zeolite.

9. The process of claim 1 wherein the catalyst is contacted with the buffered oxalic acid for a time sufficient to extract metal contaminant without dissolving the support.

10. The process of claim 8 wherein the catalyst is contacted with the buffered oxalic acid for a time sufficient to extract metal contaminant without dissolving the support.

11. The process of claim 1 wherein the buffered oxalic acid solution is a mixture of oxalic acid in an aqueous/non-aqueous solvent system.

12. The process of claim 10 wherein the buffered oxalic acid solution is a mixture of oxalic acid in an aqueous/non-aqueous solvent system.

13. The process of claim'1 wherein the metal contaminant is one or more metals selected from the group consisting of iron, vanadium, nickel, sodium, magnesium, chromium, copper, strontium, lithium, and lead.

14. The process of claim 12 wherein the metal contaminant is one or more metals selected from the group consisting of iron, vanadium, nickel, sodium, magnesium, chromium, copper, strontium, lithium, and lead.

15. The process of claim 1 wherein the contaminant is iron.

16. The process of claim 15 wherein the contaminant is iron.

17. The process of claim 1 wherein the catalyst is a reforming catalyst.

18. The process of claim 16 wherein the catalyst is a reforming catalyst.

19. The process of claim 18 wherein the oxalic acid solution is buffered with ammonium acetate.