Note: Descriptions are shown in the official language in which they were submitted.
~~~
8~37~3
This invention relates to th~ methanation of oxides
of carbon to produae methane by a procedure utilizing a novel
catalyst for this reaction. More particularly, it relates
to the utilization of certain heavy metal substituted ser-
pentines as precursors of finished catalysts whereby the
reaction between ~xides of carbon such as carbon monoxide
and carbon dioxide with hydrogen to produce methane may be
carried out under especiallly favorable process conditions, ;~
including but not limited to high conversion rates, low con-
version temperatures, long catalyst life, and relatively insen-
sitivity of the catalyst to high temperature excursions.
The production of methane from oxides of carbon by a
catalytic route has been known since 1902, and has been
employed to a certain extent in the intervening years. How-
ever, with the current shortage of all forms of energy, the
reaction has achieved great importance, inasmuch as it pro-
mises a solution to the shortage of natural gas by enabling
coal, lignite, and petroleum fractions and residues not other-
wise readily utilizable to serve as the ultimate source of
the carbon oxides which are reacted with hydrogen to form
methane (and water as the pr~ncipal secondary product).
The reaction concerned is commonly called methanation, and
a vast literature exists on the subject extending back nearly
to the commencement of the present century. Useful articles ~-
are Chapters 1 and 6 on pages 1-27 and 473-511 respectively
of Volume IV of the series entitled "Catalysis", edited by
Paul H. Emmett, New York, Reinhold, 1956. More recent reviews
showing in addition the overall process commencing with coal
or naptha, may be found in the journal Hydrocarbon Processing,
April, 1973, pages 117 through 125; and The Oil and Gas Journal,
June 25, 1973, pages 107-134. Another useful paper is Catalysis
Reviews 8 [2] 159-210 (1973).
- 2 -
... . . .. . . . .
~7~3878
The present invention deals with the final
catalyzed reaction in such composite processes to obtain
methane from low grade raw materials. It also finds
utility wherever carbon oxides are to be converred to
methane, as for example in the purification of synthesis
gas where the impurities consist of carbon monoxide or
carbon dioxide or mixtures of the two.
According to the invention there is provided in a
process wherein a carbon oxide is catalytically hydrogen-
ated so as to form methane by passing said carbon oxide
together with hydrogen over a heated catalyst, the
improvement which comprises utilizing as said catalyst an
amorphous nickel silicate resulting from the calcination
and reduction of a nickel serpentine having a chemical
composition represented by the following:
( y gM6_y_g) (RhSi4-h) 10 (H~F)8; wherein
M is Mg, Co , Fe , Cu , Mn , Zn , or mixtures thereo;
R is Al, Cr , or mixtures thereof;
Zn ' 4, Cu < 0.5, Mn ' 0.5,
g + h = 2x
O ' x ' 0.1; 0.5 ' y ' 6; and
wherein the first parenthesis shows the cations in the
octahedral layer and the second parenthesis shows the
-cations in the tetrahedral layer; and wherein from zero
to two fluoride ions may be present for a total of eight
hydroxide plus fluoride; this precursor nickel serpentine
prior to calcining being a 1:1 trioctahedral phyllo-
silicate in general having a substantially balanced
framework, from the standpoint of total charge, with no
exchangeable ions needed for neutrality, said calcination
and reduction being carried out at a temperature of about
~ - 3 -
97~3~37~
500C to about 900C and for a time sufficient to destroy
the crystallinity of the precursor nickel serpentine and
produce said amorphous nickel silicate, and reduction
being performed in the presence of hydrogen.
- 3a -
B
078~378
In general, and in the "ideal" serpentine, g ~ h ~v~, so
that the first llne o~ the formula reduces to:
~ Nl R M6 y x) (Rxsi4-x) 1 8 [la]
Since x is at most about one-fortleth of the silicon
present (on an atomic basis) and at most about one-slxtieth
o~ the combined nickel and M atom~, it will be clear that the
distrlbutlon of x as between g and h requires highly exactlng
structural determinations. On the other hand, the gross con-
tent Or Sl, Nl, M and R, as well as Or the remaining elements,
can be readily determined by standard chemical methods; and like-
wise the a, bJ and c 13ttice parameters discu~sed below can be
readily determined by standard x-ray di~rraction methods, ~o that
the correspondence or non-correspondence o~ a given preparation
or mineral speclmen to Formula [1] can eisily be determined
without separately determlning g and h, particularly since
their sum, 2x, readily rOllOw rrom the standard chemical and
x-ray investigations just mentioned.
An alternative rormulation, which upon mere inspection
will be seen to be cornpletely equivalent to the ~oregoing [1],
is that the chemical composltion corresponds to one o~ the two
end members shown below or to any composition intermediate there-
between, viz:
Y 2x 6-y_2x) (Si4) 10 (HJF)8 [lbJ
and
~ NiyM6_y) (R2XS14-2X) 10 ( 8 [lc]
all of the other restralnts shown rOr [1] of course applylng. -- !
.~ typical and indeed expected intermediate composition
between end members, ~lb] and [ lC] appe~rs herelnabove as [la~.
~ 7~7~3 :
As is known, nickel serpentines are ph~llosilicates,
that is, silicates with a laminar habit, exhibiting a basal
spacingl i.e., a c spacing, approximately 7 A. The a spacing
is about 5.3A, or a multiple thereof, while the b spacing is
a~out 9.2 or 9. 3A. Moreover, they are silicates of the so-
called 1:1 type, having one octahedral layer bonded to an
adjacent tetrahedral layer by the sharing of oxygen ions.
Moreover, the structure is trioctahedrali that is, all of the
possible sites for positive ions in the octahedral layer are
occupied, in contrast to the so-called dioctahedral structures
of some other p~yllosilicates in which only 2/3 of such sites
are filled. Furthermore, each individual 1:1 sheet is elec-
trostatically substantially neutral, with as many positive
ions as negative ions within the combined 1:1 layer structure.
This is in contrast to electrostatically unbalanced silicates,
such as zeolites and smectites, which require cations exterior
to the silicate framework in order to achieve electrostatic
neutrality. Those knowledgeable in clay chemistry will of
course recognize that some ion-exchange capacity is generated
by broken bonds at the edges of the crystallites even in
balanced layers, as evidenced for example by kaolinite and
attapulgite.
A good discussion of serpentines (including nickel
serpentines) occurs in the chapter by that name on pages
170-190 of the text "Rock-Forming Minerals", volume 3, Sheet
Silicates, by W. A. Deer et al., London, 1962. Nomenclature
for this group varies somewhat; Brindley in the text, "The
X-Ray Idè~tification and Crystal Structures of Clay Minerals",
G. Brown, Ed., London, 1961, pages 109-131, adopts the term
"nickel serpentine". French workers prefer "nickeliforous
~ - 5 -
.
- . .. . : . . ,.. - : - ~
887~
antigorite", as described, for example, on pages 180-193 of
the text "Minéralogie des Argiles", by Simonne Caillère et al.,
Paris, 1963.
- 5a -
': -
:
~078878
Nlckel serlentine~ withln the sco~e of the invention
occur in a number Or localltles throughout the world, some in
quallty and quantlty Or such a nature to be commerci31. The
nlckel ores Or New Caledonla are l~rgely nickel serpentlnes J
and indlvldual species thereof have been called nepoulte, noumeite,
~nd others. Another well-known deposlt occurs near Rlddle,
Oregon, and h~s been generally identlfled as garnlerite. Ty~lcal
analyses of the New C~ledonian and Oregonian nickel serpentines
are given in U.S. Geological Survey Bulletln No. 770, 1924J
p. 712. It wlll be re~dily understood thflt ~11 mlnerals which
m~y be termed nickel serpentines may not necess3rily have a com-
position coming wlthin the scope Or formul~ [1] set rorth herein
above; an analysis dnd inspection wlll of course settle the
point in any glven case.
In a~dltlon to naturally occuring nickel serpentines, a
vast llterature, both scienti~lc and patent, exlst~ dlscloslng
varlous methods for syntheslzlng nickel serpentines, and more-
over, for synthesizing nickel serpentines within the scope of the
invention provided that the startlng materials and processlng
condltions are such th~t ~ synthesized product colnlng wlthln the
scope Or formul3 [1~ result~.
Svme Or the r-elevant llterature of thls type comprlses
the ~ollowing ~rtlcles:
-6- -
. .
7887~3
C.R. Acad. Sci. P~ris, Serie C 264 [18] 1536-8 (1967)
Mlrtin, G.A., et al.
Sur la ~reparation et la structure de l'antlgorite et de
la montmorillonite de nickel.
Ibid., 225, 869-72 (1947)
Lon~uet, J., et al.
Synthese de silicates de nickel, magne~ium et cobalt~
presentant des structures du type kaollnite-antigorite.
Ibid., 239~ 1535-1537 ~1954)
Caillere, S., et al.
Synthese de~ quelques phyllites nickellrere~.
Ibld. J 241, 810-812 (1955)
Caillere, ~., et al.
Inrluence de la temperature . . ~ormation de l'antigorlte nickelirere.
Ibid., Serie C 267, 610-613 (1968)
Dalmon, A., et al.
Sur la preparation et la structure des silicates basiques de
cobalt et de magnesium du type talc et antlgorite.
Journal de Chemie Ph~si~ue et de Physicochimie Biolo~ique 67 (b)
1149-60 (1970)
Martin, G.A., et ~l.
Synthese du talc et de l'antigorite de nickel, etude de
leur decompositlon thermlque et de leur reduction en vue
d'obtenir des catalyseurs de nickel sur silice.
Helv. Chlm. Acta 25 1543-47 (1942)
Feitknecht, W , et al.
~ber die Bildung eines Nickel- und Kobaltslllcates mit
Schichtengitter
N3turwissen~cha~ten 39, 233-34, ~1952)
Noll, W., et al.
Synthese des Garnlerites
781!37~3
.~merican Mineralo~ist 39: 957-75, (lgs4)
ROYJ Della M,, et al.
An Experimental Study o~ the Formatinn and Pro~)erties
Or Synthetic Serpentlne~ and Related Layer Silicate Mlnerals,
Clay Minerals Bulletin, 5, ~no,30~ 272-278 (1963)
Caillere, S., et al.
Nouvelles Etudes sur la Synthese de~ Mlnerzux Argileux a ~artir de gels.
Trans. 4th Intern. Congr, Soil_Scl., Amsterdam 3, 34-37;
Franzen, P., et al.
O Synthesis Or Nlckel Hydrosilicates
Bull. Grp. ~ranc, Ar~iles 7, ~2] 21-30 (1956)
Caillere, S., et al.
Etude de Quelques Silicates Nickéli~eres Naturels et de Synthese
Contr. Mineral and Petrol. 34 84-86 (1972)
Jasmund K., et al.
Synthesisor Mg^ and Ni-Antigorite
Ibid,, 34 346 (1972)
Jasmund, K,, et al.
Synthesis Or Mg- and Ni-Antigorite: A Correctlon
Beitrage zur Mineralogie und Petrographie ~, 232-241 (1960)
Noll, l~., et al.
Ueber synthetischen Kobaltchrysotil und seine Beziehun~en usw.
Bul. Soc. Franc. Miner. Crist. 79, 408-420 (1956)
Caillere, S,, et al.
i Étude Exper~mentale du mécanisme de la rormation des antigorites
nickelifères. ,
Kolloid-Zeitschrlft ~ 11 (1958)
NO11J W., et al.
Adsorptionsvermoegen und spezi~ische Oberflaeche von Silikaten mit
roehrenfoermig gebauten Primaerkristallen.
--8--
~ 7~3878
Typical o~ the patent llterature containlng procedure~
~or gynthesizing nickel serpentines and other so-called nickel
serpentine minerals are the ~ollowlng:
u.s. 2,658,875 Schuit~ et al., November 10, 1953
53,686J341 Eberly August 22, 1972
3,686,348 Eberly August 22, 1972
3,692,700 Sa~yer et 31. September 19, 1972
3,729,429 Robson April 24, 1973
3,8~,741 Robson April 16) 1974
l3,838,041 Sawyer et ~1. September 24, 1974
., .
8~7~3
`- The precursor nickel ~erpentlnes are made by a hydro-
thermal proce~s; or they may be naturally occuring nlckel
serpentlnes corresponding to the fore~oing descriptlon, to the
extent that commercially workable deposlts thereof are avail-
5 able to those desiring to practice the invention, as alreadydiscussed. In general we prefer to syn~hesize nickel serpen-
tines ina~much as better control may then be had over the
properties o~ the flnal product, with part-lcular rererence to
such features as chemlcal composition, partlcle configuratlon,
surface area~ and the like.
When the precursor nickel serpentines are to be obtained
; by synthesis, a procedure selected ~rom the extenslve prior art
syntheses listed hereinabove may be used A general procedure
which we prefer may be carrled out as ~ollows:
In general, relatively simple ~ources of the various
elements present in the desired product and an alkali such as
sodium hydroxide are added to water. The well-homogenized
mixture i~ placed in a sealed pressure vessel, which is then
brought to a preselected temperatureJ typically 250 C. to
350 C., and maintained there for a preselected period of time,
typically one-hal~ or 60r even 72 hoursJ after whlch the ve~sel
is cooled and the contents removedJ andJ i~ desired or necessary,
washed ~ree of soluble salts and dried. Agitation durin~ the
hydrothermal processing is generally desired ln large-scale
preparations. The product i~ convenlently examined by x-ray
dlffractlon as a check on any given run.
A general ~eed formula utilizing soluble salts is as ~ollows:
yNiC12 (6-y-x)MgC12 2xAlC13 (4-x)Sio2 (12-~ 4x ~ ~ ) NaOH nH2 0
where ~ is the mols of base in excess of that necessary to pre-
cipitate all multivalent metals a~ thelr oxides or hydroxldes,and where n is approximately 200-350, preferably about 250.
--10--
_. . . . _ . . _ _ _ . . . ~
7~
The value~ o~ x ~nd y are the same a~ those set rorth
ln the nickel ~e~ent:Lne ~ormula [1] given earller in this
dlsclo~ure; we have ~ound that the 9ynthe~i~ed ~roduct~
exhlblt substantially the same ratlo of metal constltuents
; as ls present in the ~eed mixture, 90 that the values of x
and y are the sa?ne for both reed and product; and are sub~ect
to the same restralnt~ as already given in f`ormula [1~. For the
NaOH there may be sub3tituted KOH, LiOH, NHl~OH, 1/2 Ca(OH)2,
. or mlxture~ thereo~, or llke alkalizing agent.
O The foregolng feed formula shows the various metals added
a~ their chloride salts. This i8 in generdl preferred, although
other ~alts may be used, such a9 the cltrate, acetate, nitrate,
and the like
In general, ~ in formula [2] above may vary from zero to
as hlgh a~ about 20, the higher values Or ~ being practical for
weak bases such as NH40H. For strong bases a ~ value higher
than 6 19 scarcely needed.
The constituents, save for the caustic a~d a mlnor portion
o~ the added water, are conveniently placed in a suitable mixlng
'O device and homogenized, a~ter which the cau~tic is added in
solutlon with contlnued mixing, For small-scale laboratory
runs, a sllver-lined stalnless steel pressure vessel wlth a
capacity of about 15 ml may be used, and heated ln an oven
For batches of larger slze, autoclaves of sultable capaclty
~'5 may be used.
78~
\
The reed formula ~u~t glven does not recite ~luoride, whlch,
Or course, i9 present whenever this optional component 19 desired
to be present. When this 19 the case, we rind it best to add
~luoride as sodium ~luorlde to the ~eed mixture wlthout attempting
j to reduce the ~mount of caustic to compensate. Indeed, quite
in contrast to the metallic ions, not all o~ the ~luorlde ions
present in the ~eed mlxture ultimately become part of the nickel '~
serpentine lattice, 90 that it is necessary to use a con-
siderable excess Or sodium ~luoride. Up to six or eight mols
() of ~odium ~luoride per formula weight ln the ~eed may for example
' be used, which wlll still result in somewhat less than one mol
of rluorlde becoming part o~ the lattice; that iSJ slightly
less than one hydroxyl will be isomorphousIy substltuted by
fluoride. More sodium rluorlde may of course be use~, leading
to a higher degree o~ substitution.
Consideration o~ the ~ormulas ror the feed and for the
synthesized product will rnake it evident that soluble salts,
~or the most part sodium chloride, will appear as an admixture
with the product when this type o~ reed mixture 19 used. In
~0 general it is desirable to separate out the soluble salts by
thorough washing ~ith water. The washed product may then be
drled, and lf deslredJ ground.
The silica may be ~dded ln any convenlent ~ormJ such as
polysiliclc acid J which may be r~ade in accordance with United
~5 States Patent 3J649J556 to Hor~man. AlternativelyJ sodium
silicate solution mdy be u~edJ taklng into account the caustic
soda equivalent thereo~. The alumina 19 convenlently any ~inely-
dlvlded reactlve rorm thereo~J such as alumlnum hydroxlde, some-
time~ termed alumina trihydrate.
When the synthetic procedures set ~orth ln the prlor art
articles and patents cited hereinabove are employed, due conslder-
ation should Or course be given to the compositional restraints
~peci~ied in accord~nce wlth the inventlon.
7~387~
The nickel serpentine~ may also be readily ~yntheslzed by
using a feed mlxture in whlch the various component~ are added
in the form of their oxidesg basic oxid.e~ or carbonates, wlth-
out added alk~li as such. This obviate~ the nece~lty for
w~shing the product free of ~oluble salts. Examples showing
this synthetic route wlll be given hereinbelow.
1'
, .... . ...
` ` ~L071 3~
As we have explained hereinabove, ln order to ~orm our -
inventive catalysts we calcine and reduce the selected niclcel
serpentines so that their crystalllnlty iq destroyed,which may
be termed "amorphous". One must, Or course, understand that this
by no means implies that the structure is utterly ran~om. To
the contrary, the relative dispositions o~ the component ions have
been conditioned by their previous crystall:lne structure while
in the precursor serpentlne state. However, in view Or the
number Or di~erent atomic species present, it is quite beyond
present-day technology to determine precisely what those dis-
positlons are, so that the inventive catalysts can only be
characterized in terms o~ their crystalline precursors and
of the processing by way o~ calcination and reduction
It will be understood that by "amorphous" we rerer to the
essential non-appearance o~ the characteristic serpentine struc-
ture in the x-ray di~rraction pattern, although diffuse metal
oxide c~ystal patterns may appear, as ror example bunsenite,
i.e. J nickel oxide, arter calcination,and crystallites of -~
metallic nickel arter reduction.
The temperature required ror calcination can Or course be
readily determined by pilot tests; a temperature of about 700 ~.
to about 750 C. will be found adequateJ although the erfective
range varles rrom abou~ 500 C. to about 900 C. The calcined
product is then reduced by heating in a reducing, preferably
hydrogen, atmosphere at an e~rective reducing temperature, which
is ~ound to be ~rom about 500 C. to about 900 C. In both steps,
the duration should be long enough to obtain the desired degree
of reduction. Generally, three hours at 705 C. is adequate for
calcination, and five or more hours at 650 C. ror reduction, the
hydrogenJ pre~erably at a minimum Or one atmosphere. It will be
appreciated that these temperatures and times are given by way of
-14-
~L~'78~
.,
xample and not by way o~ limltation, Simple preliminary
tests well known to those in the x ray di~raction and
catalytic arts wlll provide satls~actory processlng condi-
tions ~or any given case, The calclnation and reduction
can be conducted concurrently by heatlng the serpentine at
a temperature above about 500 C. in the presence Or hydro-
gen, In generalJ howeverJ the serpentine ls less easily re~
duced than the amorphous calcined serpentine and longer re-
ductlon times are necessary when the serpentine is not cal-
cined before reduction.
1~8~7~3
We now give some examples showing the practice Or our
lnvention:
Example 1
Four mols Or polysilicic acldl made as described in
U.S Patent 3~6l~9J556~ and six mols Or nickel carbonate
(materials assaying 80.77~ NiC03 was used) were made into a
slurry with water at 15~ total solids by weight with a high-
speed mixer. The slurry was placed in an autoclave a~d maln-
tained at 300 C. ~or rour hours, The apparatus was then
cooled, and the nlckel serpentine recovered. It had the fol- :
i lowing ~ormula:
.
(Ni6 ~ (Si~ ) 10 (OH),8
Separate portions of the product were reslurried w~th ; :
various clay binders in the proportion o~ 80~ by weight o~
the nickel serpentine and 20% o~ the selected clay, using
water and a laboratory mixer, and then dried, calcined at
705 C ror three hours, and ground to 30/60 mesh. One por-
tion was also used without admixture, as the pure nickel ser-
pentine. The samples thus prepared were given the rollowing
designatlons:
Sam~le Clay Binder
1 A None
1 B kaolinite, from Georgia, U.S.A.
1 C saponlte, made as described in
U.S. 3,855,147 ~a = OJ X ~ 1.43~ Y ~ O)
1 D sepiolite, rrom Nevada, U.S.A.
1 E hectorite, rrom Cali~ornia, U.S.A.
1 F attapulgite, ~rom Georgia, U.S.A.
1 G montmorillonite, from Wyoming3 U.S.A.
-16-
,:
~78~
The samples prepared as descrlbed above were then tested
for their activity aa methanation catalysts. 8.4 cc Or each
sample were placed in the reactor tube and reduced by heating
in a hydrogen atmosphere for sixteen hours at 650 C., the
hydrogen being passed through the sample (having a mesh size
of passing 30 and retained on 60 U,S. Standard mesh) at a rate
o~ 40 cc per minute. Subsequently, commencing at a temperature
of about 200 C , a mixture o~ approximately 20 mol ~ carbon
monoxide and 80 mol ~ hydrogen was passed through the catalyst
'0 charge in the reactor, the latter being provided with sufricient
instrumentation to determine temperatures, flow rates, and
input and output gas compositions, and thus the percent conver-
sion Or the carbon oxide to methane. The temperature was
raised in steps at approximately thirty-minute intervals,
to a maximum in general Or about 600 C. The input gas hourly
space velocity (the volume Or gas per volume Or catalyst charge,
per hour, at standard temperature and pressure) was 1500.
After the sequence of methanation tests Just described
had been completed ror a given sample, the catalyst charge
'0 was again reduced in hydrogen overnight at 650 C and then ror
three hours at 850 C. A~ter cooling to 200 C., a second series
Or methanation tests were commenced and carried out as berore,
ending usually at about 600 C.
Results o~ the tests are given below in Table I, while
'5 some typical gas composition data ror some Or the runs are l~
given ln Table II 1-
-17-
7~387l~
Table 1
.
Sample Gm Ni ln Minimum Tem~erature~ de~ree C,~:
8.4 cc ror 100~ C0 conversion., ~or ~ 90~ C0 conv.
Charge Cataly~t reduced at Catalyst reduced
650 C. . at 850 C.
1 A 3.16 235 470 ,:
1 B 2.61 220 250
1 C 2~32 202 ~ 315
1 D 1.90 202 310
1~ E 2.7~ 243 45 ;
1 F 2.11 240 320 ,.
1 G 2.38 220 300
'. : ` . '
37~87~
Table II
Sample ~C, of Exit ga~ composition~ mol percent Percent C0
No. run C0 H2 C~LI C2 converslon_
1 A * 235 0.02 50.5 49.1 0.34 99.97
* 393 0 50.7 49.3 0 100
* 590 0.15 60.4 38.5 ~.o 99.6
** 470 1.5 64.5 26.4 7.6 95.8 ,
** 600 0.3 61.4 36.5 1.9 , 99.3
1 B * 220 0.06 55.21 42.232.49 99.9
* 235 0 55.1 1~4.70.21 100
* 500 0 57.7 42.3 0.02 100
* 600 0.31 53.9 44.1 1.7 99.3
** 250 0 64.3 3005 5.2 100
** 600 0 58.8 41.1 0.1 100
1 C * 202 0 60.8 37.~ 1.5 100
* 398 0 59.6 40.4 0 100
* 595 0 67.2 32.8 0.05 100
** 315 o 57.4 39.6 3.0 100 ,
** 585 0 59.3 40.2 0.52 100
1 D * 202 0 49.6 49.2 1.2 100
* 495 0 46.4 53.6 0.03 100
* 569 0.27 54.6 43.5 1.7 99 - 4
** 310 0 60.8 33.1 6.1 100
** 580 0 53,8 41~ ~ 7 1.6 100
-19- :
.. . . . . .
~7i 38~
.
Table II (continuedl
Sample . C. o~ Exit gas composition, mol ~ Percent C0
No. run C0 H2 CH4 C02 conver~lon
1 E * 243 0.14 51.61,7.0 1.2 99-7
* 410 0 51.348.7 0 100
* 610 0 56.642.9 0.6 100
** 405 0 50.548.3 1.2 100
** 500 0 48.351.7 0 100
*~ 600 59~140.7 0,2 100
) , ' 1 F * 230 0.9 52.843,3 3.0 98,2
* 2L~5 0 49.247.8 3. lO0
* 600 0 49.349,1 1.6 100
** 320 59.834.3 5.9 100
** S20 0 47.652.2 0.2 100
** 610 0 52.546.6 1,0 100
* Runs on catalyst reduced at 650 C.
** Runs on catalyst reduced at 650 C, overnight and at 850 C.
~or three hours.
-20-
~ 7~37~3 :
Pne may see ~rom Table~ I and II that thls niclcel ser-
pentlne is an excellent methanat~on catalyst, givlng substan-
tially complete conversion o~ carbon monoxlde at a low tem-
perature, vlz., 235 C Moreover, even a~ter being sub~ected
j to the hlgh temperature Or 850 C. lt is stil~ a good catalyst
giving nearly com~lete converslon at 470 C. Still ~urther,
it operates well over the range up to about 600 C., whether
lt was reduced at ~50 ~. or 850 C The criterion is whether
the reaction between the hydrogen and the carbon monoxide has
been catalyzed to give essentially equilibrium conversion o~
the starting materials, which is herein termed complete, or
100%, conversion. The thermodynamics Or the reaction are
more ~avorable at lower temperatures. In addition, low oper-
ating temperatures have obvious advantages ln equipment sim-
pli~ication and longer expected life. However, the reactionshould not be conducted below about 200 C because o~ the
formatlon of nickel carbonyl at lower temperatures.
Table II illustrates an important aspect o~ the lnventlon,
in accordance with which the nickel serpentlne is lntlmately
~?0 admixed wlth up to about lts weight Or a water dispersible
clay mineral, prererably by wet mixing and subsequent drying.
Suitable such clay mlnerals include montmorillonlte, bèidelllte,
kaolinite, halloysite, saponlte, se~iollte, hectorite, atta-
pulglte, lllite, and others. To be ~ure~ the clay, particularly
when incorporated in the ~ashlon descrlbed, imparts strength
and mechanical lntegrity to the catalyst particles, but it is
surprising and entirely unex~ected that the clay addition causes
the catalyst to give complete conversion at lower temperatures
than otherwise, even when the clay-bearlng nickel serpentine
catalyst i9 reduced at the hlgh temperature of 850 C. The
latter ls of practlcal lmportance because momen-
-2l-
- .... . . . , : . .
~L~78~37~
tary higll excursions Or temperature are well-nigh unavoidable
in plant operation, and a good catalyst should remain undamaged
when so treated.
Thus, considerlng Tables I and II, it may be seen that
; 20~ by weight o~ kaollnite lowers the 100~ conversion temper-
ature rrom 235 C. to 220 C., and even more lmportantly lowers
the complete converslon temperature a~ter the 850 C. treat-
ment rrom 470 C. to 250 C. Similar beneficial e~ects may
be observed for the other four clays used as binders in the
runs Or 1 B through 1 F.
or course, more than one clay mineral may be used in a
single preparation. For example, the nickel ser~entine prlor
to calcination may be lntimately admixed with u~ to its own
welght of a mixture of equal parts Or kaolinite and montmorillonite;
or a mixture of equal parts Or ~aponite and attapulgite, and
so rorth, all ~re~erably by wet mixlng.
The tables are also exemplary Or the circumstance that
the methanation resultlng rrom the hydrogenation should be
carrled out at at least 200 C, (to avold nickel carbonyl for-
mation) but in any case at a temperature high enough to
e~fect substantial methanation Or the carbon oxide or carbon
oxides involved, The latter temperature varle~ from one cal-
cined and reduced nickel serpentine to anotherJ as is evldent
~rom the test results given hereinabove, and elsewhere herein.
-2?-
8~
Example 2
The tests just as deacribed were carrled out on a ~urtherseries o~ mlxtures of the same nlckel serFentlne (designated
as 2-A) wlth dirrerent proportion~ Or montmorillonite, in the
form Or Wyoming bentonite rully converted to the sodium form
and rreed Or dross by supercentrirugingJ ion exchanging, and
spray drying. These catalysts were calcined at 705 C ror
three hours, ground to 30/60 mesh, and reduced overnight, as
berore, at 650 C.. After methanation testing, the catalysts
were reduced at 650 C. overnight and at 900 C. for three
hours berore additional methanation testing was undertaken. ~-
The results rollow: :
Table III :
Methanation Activity: . ::
~ by weight Minimum C. ~or 90~ C0 Conversion
Sample Montmorillonite * **
r:
2 A 0 . 225 ***
2 B 5 210 250
2 ~ 10 220 300
2 D 20 215 250
2 E 25 215 215
2 F 30 235 250
2 G ` 50 235 ***
* Catalyst reduced at 650 C. overnight
** Catalyst reduced at 650 C. overnight and then ~or three
hours at 900 C.
*** Maximum conversion c 90% occured at 600 C
',
-'3-
788 ~1~
Table III shows that montmoril~onlte shares the pro~er-
ties of the other c~ays listed ror sam~es lB through lF of
reducing the erfectlve methanation tem~)erature, both when
reduced at 650 C. and at 900 C. It will be recalled that
sam~les lB through lF contained 20~ clay and 80% nickel ser-
~entlne. The results shown in Table III ~or samples 2B through
2G exhibit the e~ect Or varying c~ay content over a wide range.
At 50~ montmorillonite (sample 2G) the activity o~ the catalyst
reduced at 900 C. is ~oor, but the 235 C. ~igure ~o~ the cat-
alyst reduced at 650 C. is remarkable, considering that this
sam~le has only hal~ as much of the nickel ser~entine (before
calcinlng) as sample 2A. It would a~pear that some synergistic
action occurs between the clay and the nickel serpentlne to
account for this extraordinary behavior, slnce clay minerals
by themselves have substantially no methanatlon activity. ;
.
-2!l -
37~3
Example 3
The synthesls descrlbed in Example 1 was carried out
uslng, in additlon to the polysilicic acid and nickel carbon- -
ate, magnesium oxlde and alumlna. The molar ratios of the
several components were ln the same proportion as the nickel
serpentine formed, viz.:
4.425 gl.5Alg) (S13 925Alh) Olo(OH~8J
where g ~ h = 2x ~ 0.15.
Methanation tests and catalyst preparation as de-
scrlbed ln Example 2 were carried out on the sample, with
the rollowing results.
Table IV
'
C. of Exit gas composi.tion, mol percent Percent C0
run C0 H2 CH4 C02 Conversion
220 * 0.1 57.9 39.0 3.0 99.7 ~::
300 * 0.0 50.8 48.8 0.4 100.
I~oo * 0.0 48 9 50.9 0.2 100.
500 * 0.0 50 6 48.7 o.8 100 .
600 * 1!0 56.6 39.~ 3.3 97.6
300 *~ 0.2 56.6 39.9 3.L~ 99-7
L~oo ** o~ o 55.0 42.3 2.6 100
500 ** 0.1 53.3 44.9 1 7 99 9
600 ** 1 2 57.3 37.8 3.7 97-3
* Runs on catalyst reduced at 650 C.
** Runs on catalyst reduced at 650 C. ove.rnight and r'
then at 900 C. for three hours.
It will be seen t.hat thls nickel serpentine wa~ like-
wise the precursor of an e~cellent methanation catalyst.
-25 -
7887~
Example 4
A series of nicke~ serpentines were Frepared by
hydrothermal synthesls rOr four hours at 300 C , with
varying amounts of nlckel per unit cell. l'he general
procedure was that o~ Example 1. Table V which rollows,
gives the grams Or the starting materlals, all batches
belng made up to 15% total solids. Also shown are the
expected unit cell compositions o~ the products.
Table V
Sample S102 (1) NiC03 (2) MgO (3) Unit Cell Composltlon
No grams grams grams
3A 1200 4060 100 (Ni5 5Mgo 5) Si4lO(OH)8
3B 82 l~ 236.9 14~4 (NisMg) Si401o(0H)g
3C 135.6 373.2 35.2 (N14 5Mgl 5) Si4lO(OH)8
3D 145.8 268.3 76.7 ~Ni3Mg3) Si~Olo(O )8
3E 158 4 145.6 124.5 (Nil 5Mg4 5) silllo(H)8
(1) As polysilicic acid
(2) 80 77~ actlve
(3) 96~ active
Berore calcination and reduction, all samples ~ere
admixed by wet mixing as before with montmorillonite as de-
scribed in Example 1J in the proportion of 5~ clay and 95
nickel serpentine.
The methanation catalytic actlvity and catalyst pre-
paration were as in Example 2, with the following result~:
-26~
78~371~
Table VI
Methanation Activity:
Minimum C. for ~ 90~ C0 Conversion r
* **
Sample
3A 210 300
3B 210 250
3C 2l~0 250 F
3D 2l~0 2L~o
3E 260 500
'
* Catalyst reduced at 650 C overnight.
** Catalyst reduced at 650 C. overnight
and at 900 C ~or three hours.
.. . .
As may be seen from Table VI, all samples performed
well, although sample 3E su~rered somewhat ~rom relatively
low nickel content. The optimum content Or nickel ~rom a
practical standpoint of course varies with the price of nickel
at the time in question, so that one must balance cost against
effectiveness. In times of high nickel prices, sample 3E
would be considered satisractory, while during times of low-
nickel prices, one would naturally choose a nickel serpentine
more along the lines Or sample 3A, 3B, 3C, or 3D
7~
Example 5
So~e of the nickel serpentines described ln Example 4
were tested for methanation activity, arter formlng into
catalysts by calcining at 705 C. for three hours and hydro-
gen reduction at 650 C, overnightJ for the conversion of
carbon dioxlde into methane. This reaction is more difficult
to catalyze than that of carbon monoxide, as is generally recog- ;
nizedJ although it does not have the same commercial importance
as the methanation o~ the latter.
The nickel serpentines were mixed with 5~ montmorillonite
as already described, and the mesh slze and catalyst charge
were likewise the same as in,the earlier examples. The reed
gas was 15 mol % carbon dioxide and 85 mol ~ hydrogen.
Results are shown in Table VII as follows:
. .. .
- 2~)
.f
~ 7~7~
, .
Table VII
Sample c,Or Exlt Gas Composition, Mol ,q6 _ Percent C02 L
No. run C0 ~l2 CH4 C2 Conver~lon
3 C 280 o 66.9 27.7 5.4 83.6
300 0 6?.1 35.3 2. 6 93.1
400 0 56.9 42.4 0.7 98.4
500 o.4 61.9 34.6 3.1 91.9 ;~
600 2.3 71.2 21.6 4.9 82.9
3 D 300 0 82.0 14. 8 3.2 82.2
325 0 80.3 19.2 0.5 97.6
400 o 81.1 18.9 0 100 .
500 0.1 77.6 22~0 0.3 98.6
600 0.5 79.7 16.8 2.0 90.1
3 E 320 0 77.6 16.0 6.1~ 71.4
340 0 76.3 20.1 3.6 8500
400 0 70.1 29.4 0.5 98.4
500 0.2 72.7 25.9 1.2 95.5
600 2.1 80.2 ~4.6 3.1 84.6
The excellent catalytic e~rectiveness for carbon dioxlde
methanation may be seen rrom Table VII. Substantial catalytic ..
actlvlty set in at around 350 C.
.
-2'~-
1~'7~
.
Example 6
A natural nickel ~erpentine rrom Riddle, Oregon, was
tested as a precursor for a methanation catalyst for carbon
monoxide. The analy~is Or the materlal drled at 110 C wa~ :
as rOllOws:
Table VIII
Constltuent Wei~ht Percent
N10
Fe203 (total) 3.59
. Mn23 0 04
Cr~203 O. 01 ,
Ti2 0 003
CaO 0.05
K20 0 04
S102 44 51
A123 o, 42
MgO 29.97
H20 12.63
Total:100.22
X-ray di~fraction analysls showed a typical nickel ~erpen-
tine, with a spacing of approxlmately 7 A
. The mlneral, which had been ground to pas~ 200 mesh, was
calcined in air at 705 C. ror three hoursJ and then placed
in the reactor, where it was reduced with hydro~en ror rourteen
hour~ at 650 C
-30-
7~38~
. :
Uslng a gas feed Or 20 mol percent carbon monoxide and
80 mol percent hydrogen, the ~ollowlng percent conver~ion~
were obtained:
Table IX
Reactor Temperature, C. Percent C0 Conversion
_ _
300 58
350 90
400 100
450 99
"
These results are quite good, especlally considering the
relatively low nlckel content o~ the sample, whlch was about
0.78 nickel atoms per unit cell, corresponding to .y = 0.67
ln Formula ~1].
-3L-
~8~7~
, . :
The roregolng disclosure illustrates the manner of carrying
out the invention, and shows many o~ the advantages thereof.
It will be evident to those slllled in the art that in many
methanation installations on a commercial scale, the oxides Or
carbon to be converted to methane will be prese~t in a mixture
Or various components Thus, carbon monoxide and carbon dioxide
may both be present; and depending upon the source Or the carbon
oxides and Or the hydrogen, more or less other gases such as
nitrogen, water vapor and the like may be present. These will
in general ofrer no bar to the successful carrying out Or the
invention.
In Equation [1], and in the dlscussion of the reed for- -
mula whereby the precursor nickel serpentines used in the in-
vention may be ~ormed, rluoride is shown as an optional com-
ponent. In general this ~ay be omitted without appreciably
altering the bahavior of the calcined catalyst. However it
i8 sometimes an aid ln crystallization ~rom the starting ~eed
mixture, and its optional inclusion has been set ~orth herein
for that reason.
'0 ~ngstrom units (one-tenth nanometer) have been abbreviated
"A" ln accordance with common usage, ln this speclfication.
10~7~ 7~3
. ~ ~
It will be understood that while we have explained the
inventlon wlth the ald Or .specl~ic examples, nevertheless
considerable varlation is possible in choice of raw materials,
proportlons, processlng conditions, and the like, wlthin the
broad scope Or the inventlon as set ~orth in the claims whlch
rollow. Thus, ror example, our inventive catalyst may be
used simultaneously with other catalytic materlals, so as to
sult particular eonditions and circumstances. Further, the
calcination and reducti.on may be carried out as separate or
overlapping steps.
, :
: ....
