CA2088150A1

CA2088150A1 - Catalysts and getter systems

Info

Publication number: CA2088150A1
Application number: CA002088150A
Authority: CA
Inventors: Alan Edward Heywood
Original assignee: Individual
Current assignee: PGP Industries Inc
Priority date: 1990-07-31
Filing date: 1991-07-30
Publication date: 1992-02-01
Also published as: FI930374A0; AU8326491A; AU661971B2; EP0544710A1; WO1992002301A1; GB9016787D0; IE912688A1; JPH06503744A; FI930374A; ZA916028B

Abstract

2088150 9202301 PCTABS00010 Knitted precious metal textiles such as wire gauzes, and methods of making the same, are disclosed. The textiles are suitable for use in catalysis, and are especially useful for the catalytic processing of ammonia. Particularly preferred knitting stitches are tricot, jacquard and raschel. Rotary or circular knitting machines may be used, but warp knitting machines are preferred for most applications. Advantageous products produced by the knitting process are described. These products include layers of non-uniform thickness, and uncut products having non-parallel side edges, such as circles.

Description

WO 9~ 2301 2 ~ 3. ~ {~ p ~r,G~l/0l293 , i ~ , 1 .. . .
I~PROVE~SENTS IN OR R~LA ING TO
CATA.~YSTS AND GETTER SYSTEMS

This invention relates to catalysts and to getter systems for catalysts. The invention relatas par-tlcularly, ~hough not S exclusively, to catalvsts and getter systems that are suita~le for ammonia oxidatien syscems. ~xamples are catalysts used in the production OL ~'' t.ic ~cid, hydrocyanic acid and other amrnonia oxldatioTl ?roduc Ls ~

The o~ida~ion o:. z~mon~ Y';3) ~o fo~ s-har userul produc~s, 10 such as nitric acld (-hhO3) or hydrogen cyanide ('nCN), is generall~ effec~ed, in industry, b-~ a c2tal-~is ~rocess. This process invclves one o_ mo_e p-aeious metal oxidation catalysts. The most widely used catalyst is platinum, either alone or in combination with other platinum group metals, such 15 as rhodium and palladium.

For much of this century, catalysts for reaction~i such as the oxidation of ammonia have been in the fonm oi wov~n g~uze of, for example, a platinum group metal (pgm) alloy. Getter systems for catalysts also rely upon woven sauze. The gauze 20 is usually m2de u~ of ~ires, but othsr elong2te elemen~s such as strips may also be used; in this specification, the term ~wire~ is in~ended ~o encompai~is such o$her elonga~e elementsO

Iypically, in use in a reactor chamber, th~ wire gauze is . .:
~ supported in the path of reacting gases or fluids, which flow 25 thorough the gau~e and contact the ca~alytic material under controlled conditions.

Weaving of catalytic gauzes is commonly performed on a loom, in a process very similar to the weaving of textile cloths, i.e. with individual warp and wef~ strands of wire. The mesh 30 thereby obtained is usually regular, with rectangular : ~ normally square~ intersticas. Yarlants to this plain wea~e WO92/02301 PCT/~B91/n1293 ~3~ 2 -include herringbone or twilled weave. The nature of the weaving process also means that the gauze itself is rectangular in shape when it comes off the loom.

Modern gauzes commonly feature 80 meshes per square inch (10~4 5 per cm2), using 0.003 inch diameter wire ~0.076 mm). Other wire diameters in common use are 0.0024 inches (0.060 mm), O.0027 inches (0.085 mm), 0.0036 inches (O.090 mm), and 0.1 mm .

The thickness of the wire and the size of the mesh is 10 impor~ant, because the reaction gases must pass through ~he wire gauze at a reasonably rapid rate tha~ is consistent wi~h the e'ficiancy of the catalyst, but without ~xcessi~J~ ~ac'~
pressure. The wire thicXness is significant in ~wo r~spec~s:
firstly, the thicker a wire, the larger its surface area and 15 secondly, the thicker a wire, the longer it can last undsr conditions of volatilisation (evaporation).

~earing in mind that the oxidation of a~unonia is a heat~
intensive process involving an exothermic r action, the gauze must be able to withstand the stress of very high tempex~ture 20 and flowing reactive gases without excessive degradatiGn, and without excessive loss of catalytic or reaction eficiency.

Known, woven gauzes suffer from a number of disadvantages.
Weaving is a slow process even if relatively fast 'rapier' or ~projectile' looms are used. This results in production 25 delays, and represents a high production cost in terms of area of catalytic gauze produced per hour. Also, where the gauze is made of precious metals and is therefore of subs~antial value, the slowness of weaving ties up expensi~e capital assets for an undesirably long period of time. This 30 phenGmenon, known as "metal lockup~, means that large reserves of precious metal lay idle, and are isolated from world markets at great expense, while awaiting processing in~o ~oven gauzes.

WO92/02301 ~ PCT/GB91/01293 The slowness o~ eaving is made all the worse by the need for laborious setti ~ up procedures whene~er a new gauze is to be produced, or even when a wire of diffarent composition or diameter ie to be introduced into the gauze~ It is also 5 difficult to vary wire density. In short, the weaving process lacks flexibility.

Woven gauzes themsel~es suffer from problems. For example, because woven gauzes are generally of rectangular shape when produced, they must be woven 'over si~e and then cut do~m i~
10 the desired final shape is circular or otherwise non-rectangular. The of fcuts constitute scrap and, consequently, represent unde~ rable waste. The cutting process adds further ~roduction time.
,:
Woven gauze is also prone to warp under thermal stress during 15 exothermic reactions, forming a tortoise-shell pattern of creases, ridg~s or wrinkles. ~hese ridges can cau~e undesirable side reactions in some processes. In hydrogen cyanide reactors, for example, they can gi~e rise to carbon deposits be ause the effe~tive bed depth, and hence residence 20 time of reaction gases and by-products, is increased at the site of the ridges. Carbon may then be deposited on the surface of the catalyst, thus obscuriny the catalytic surface much to the detriment of catalytic efficiency. The carbon can even combine with the catalytic material, causing 25 em~rittlement and thus increasing the risk of mechanical failure~ In nitric acid reactors such ridges also cause local areas of g.reater effective catalytic depth resulting in different rates of reaction and fficiency.

The layers in a catalyst pack may also weld toge~her by 30 dif~usion welding at the site of any ridges, further limiting catalytic efficiency due to reduce surface area. Welding also restricts the ~reedom of movement of the layers with respect to one another, further worsening the wrinkling problem.

WO92/02301 q~ ~ PCT/CB9i/012~3 woven gauzes also suffer from an uneven di.stxibu~ion of mechanical strength through~ut tha gauze, which is an inherent consequence of the w~avina process and its use of perpendicular warp and weft wires. woven gauzes are also known to experience stress failur^- in the reactor, due to the pressures of flo-~-ing gas~s. Thermal expansion OI the gauze in the reactor also can lead to serlous Prohlems. Conventional gauzes deteriorate a~ the nigh temperatur2s of many industrial oxidation re2ctlons.

10 Precious me~al gauzes concinue ~o flourish deispite their disadvantages, ~ecal-s~ of -~hei r r~la~ 2 simplicity and ease of manufacture, and Decause tney are ~eli known as reliable, reasonably efficien~, and r~sonablv economic industrial catalysts. Nonethel-s~, ef ^r_5 to find i~proved catalysts and 15 catalyst supports continue, with the work of practitioners focused not so much on subs~itutes or replacements for precious metal gauzes, but on new catalyst support materials that are not made of precious metal, and which are designed to be used in combination with co~ventional precious metal 20 gauzes, or with other known ca~alys~s. Th~se suppor~s are intended to r~lie~e some of the stresses on conventional gauæes, and purport to. improve reaction efficiency in some cases.

It is against this background that the present invention has `
25 been devised. In a broad sense, the invention resides in the concept of producing pre~ious metal sauzes by a knitting - process. The invention also resides in knitted precious metal gauzes per se. Thus, the present invention provides knitted precious metAl textiles that are useful as catalysts and/or 30 as getter materials for catalyst recovery. These knitted metal textiles are particularly useful in ammonia oxidation processes, such as the production of HNO3 or HCN.

Knit~ing machines are generally much quicXer than looms in terms of area of gauze produced per ~our, once in steady WO 92/0~301 PCr/CB91/01293 8~0 production. Knitted precious metal textiles can be made as much as ten times faster than metal gauzes can be woven by conventional methods Thus, the present invention increases the speed at which catalyst and getter ma~erials can be 5 produced. In this way, the knitting process of the invention, and the resulting kniL-ted textile products, overcome or at least mitigate the economic disad~antages ol ~eaving, and in particular o~viate pro~lems of me~al loc~up and the high investmen~ costs tradition211y associ~tedTi~ th the production 10 of precious metal c~ s-r a~ gec-er ga~es.

Not only is ~nit-tiny 2 faSI~- p-OC25~ ~nan w2aviilg, but knitting machines can be set up and put into production much more quic~ly than loom5 In other riords the set-up time for knitting wir~, into te~t lc5 i_ ~I'ach shcrta than the set-up 15 time of the looms needed to weave metal gauzes. This reduces equipment and start-up costs for the production of catalysts and gotter systems.

A further advantage of the invention is that, in general, knitting machines are much more flexible than looms, ~ing 20 able to cater readily ior changes in the constituent wires simply by changing the wires that are fed to the knitting machine. Unlike a loom, which must ~ tediously rethreaded over a :3rge area when wires are chanyed, a ~nit~ing machine need only be supplied with the new wire instead of the old 25 wire. Similarlyr difrerent wires (e.g. of different alloys or diameters) can be much more readily combined in one final ~ knitted textile than can easily be combined on a weaving loom.

In general, knitting machines are able readily to produce gauzes of varying characteristics. In particular, Xnitting 30 machines can produ~e gauzes of varying shapes such as circles and hexagons, thereby minimisin~ offcut scrap which currently can approach 25~ of ~he gross ~o~en area.
` ' It is also possible substantially ~o increase the wire density ~> ~J ~ ,3 !3 Pcr/Gssl /012~3 of the gau e simply by feeding ~wo or more wires into the knitting machine in parallel. Knit~ing permits the use of more and thinner wires in a knitted textile, in comparison with a woven gauze. By means of the invention, one, two, or more 5 strands of wire can be kni~ted simultaneously. Thus, the density of the knit and the number of wires can be controlled and changed in production. This is particularly advantageo~s in getter systems, which often use more and thinnex wires ~han catalyst systems. These kinds of adjustments during produc.ion 10 are much more difficult if not impossible ~hen ~eaving a gau~e on a loom.

The invention also allows the interstices in the textile ~o be of different shapes, depending on the knitting stitch.
Conventional weaYing provides rectangular (square or oblong~
15 interstices, which limits the effective catalytie or getter surface area in comparison ~ith the shapes and flow patters that can be provided by using different knitting stitches. The knots created by knitting also provide increased wire density per vslume, in comparison with woven goods tha~ do no~ have 20 knots. This can also provide increased catalyst or getter surface arPas for a given apparent area.

A further advantage of knitting is that the resulting gauze is typically more open and flexible vr pliable than a corresponding woven gauze, while being more resistant to 25 breakage under stress. As a result of this flexibility, a knitted gauze is less likely than a woven gauze to warp into ridges under thermal stress. In particular, the textiles of the inven~ion have a close knit stxucture, rather than a conventional loose weave, and can more readily accommodate 30 thermal expansion without forming the tortoise shell ridges that are often seen in woven gauzes. T~is reduces the pr~blem of side reactions and carbon build up observed in HCN reactors that use conventional woven gauze.

The surprising advantages of knitted textiles in high-. .

W~92/0~30~ P~T/~s9~ 293 temperature environments are emphasised by the very high temperatures that prevail in hydrogen cyanide (HCN) production, in which HCN is synthesised from 2mmonia and methane in the 'Andrussow' process. Thus, a specific aspect 5 of thi.s invention relates to the use of a knitted textile as a catalyst in the synthesis of hydrogen cyanide from ammonia and methane.

The invention provides knitted texkiles made from precious metals and alloys thereof IrPferred to in this specification lO simply as 'precious metals ). In particula~, wires of platinum, rhodium, palladium and combinations thereof can be used, in proportions known to be useful for catalyst or getter applicatlons. Whilst the inv~ntion contemplates any knitting stitch, ~ire textiles produced on a x~tary knitting machine 15 or a warp knitting machine have been found particularly advantageous. Tricot stitching is especially suitable, both in ease and speed of production and in the efficiency of the final product. The tricot knit provides many large ~no~s currounding large holes, ~hich result~ in an effective 20 distribution of cataly~t or getter throughout the fabric, while permitting the reaction gases ~o pass through the fabric without back pressure problems.

Raschel or jacquard knitting techniques are also useful, as they allow greater density and unit weight, and can produce 25 a gauz~ of greater depth.

The knitted textiles of the invention can al50 be made in any desired size, based on the capacity of existing knitting machines and on the catalyst and getter support sizes commonly used in the industry. When a rotary knitting machine is used, 30 tubes of up to 30 inches diameter can be m~de, with a lO inch diameter being the most common. The knitted tube can be f lattened to provide a two-layer catalyst or getter that is up to 47 inches wide, or it can be slit to provide a ~ingle-layer catalyst or getter that is up to 94 inches wide, or two WO92/0230l ~ PCT/GB9i/0l293 single-layer pieces that are each up to 47 inches wide. When a warp knitting machines is used, a single-l~yer textile of up to 200 inches wlde can be ~ade.

Of the two types of knitting machine proposed for ~he purpose 5 of knitting gauze, namely the rotary type and tne warp type, the warp machine is currentlY pre~errod. ~his is because the rotary macnine produces a product ~hicn has to be further processed (e.g~ by slittlng o 1attenin~) to produce a flat gauze. These addltional .~an~;^acturiils c-oe-acions increase 10 production time and pro~uco~ion C051. i~oro seriously, existing rotary .kni~Ling machin~s -re of a si~o '~l~ich is inc2pa~1e of producing a seamless gauze large enough ~o suit all reactors.
As mentioned abo~7e, the l~rgest l~no-~-n rota~y machine can produce a tubul~- product o F 3 0 ~ ~ O . 7 5 ~ m ) diameter ~hich, 15 when slit and flattened out, forms a gauze approximately 94"
~2.39 m) wide. In contrast, existing warp knitting machines are capable of producing a seamles5 flat gauze up to 200"
~5.08 m) wide - enough for the l~rgest known reac~ors. Also, warp knitting machines can produce a variety of knits, 20 stitches or mesh $ypes including jacquard, raschel and tricotO
As mentioned abova, these stitches allow a large amount of catalytic or getter material to be incorporated into a given catalytic or catchment lay9r, but wi~hout r~stricting the interstices of the mesh so far as to create an excessive 25 pressure drop when the mesh is in use.

It is en~isaged that the Lnvention will give par~icular ~ benefit where the knitted wires are of intrinsically catalytic material. Thi~ is because certain reactions, parti~larly the oxidation o~ ammonia, cause a substantial los~ o~ catalytic 30 material through volatilisation. Thus, if the catalyst is not intrinsically catalytic but is merely composed of a ca~alytic layer coated onto a non-catalytic substance (e.g. a platinum- ~ :
coated base metal wire), th.is loss will, ~uite quic~ly, erode the coating until the non-catalytic surface of the substrate is exposed. Clearly, the reac~ion ~ill then cease. On the .

2 ~ ~ ~1 5 Q PCT/~B91/01293 _ 9 o~her hand, catalysts of intrinsically ca~aly~ic ma~erial will continue to present a catalytic surface during erosionl thereby sustaining the ro~ction until, even~ually, they erode away entirely. In other words, lntrinsically catalytic 5 material maximises re~ctive de~th so that, for example, a Rh-P~ wire of 0.003 (0.0~6~ mm) diarneter provides an effective depth of reacti~e or cat21ytis matori21 of 0.0015 (0.0381 mm). Providing a Rh-Pt coating of simllar depth on an autocatal~st o~ e~or. on a ~ire T./o~ld ~ inefficient and 10 costl~.

The nature o. th2 catai-~Lie ri~_aS themsel~ies depends upon the application envisaged for the catalyst. Usually, the wires will be of pl~.tin~ ~rou~ met21 ~g~.~ s~h as platlnum or palladium, ox of a ps-,~lloy wiLh Lotal p~-~ content in excess 15 of 90%. Such a composition can bê used in many reactions, for example. the oxidation of ammonia to form oxides of nitrogen or to form hydrogen cyanide; the oxidation of carbonaceous ma~erial such as carbon monoxide, hydrocarbons or polychlorinat2d biphenyl; and the reduction of oxides of 20 nitrogen or of oxides of sulphur. In other applications such as the oxidation of methanol, the wires will be of silver or of a silver alloy with silver content in excess of 75%.

Experience teaches that, in general, p~m alloys containing platinum or palladium in excess of 80% en3Oy advantageous 25 properties. Examples are 90~ platinum with lO~i rhodium, 90~
platinum with 5% rhodium and 5~i palladium, and 85% platinum ~ with the remainder palladium and rhodium. It has also been found that a pg~ alloy of 60% platinum with the remainder palladium and rhodium is effective.

30 Similarly, t'ne knitted wires can be of any suitable size, examples being in the range 0.05 mm to 0.5 mm, and jpreferably : in the range 0. 055 mm to O.l mm. The ~2i~ht of the wire can be in the range of, say, ~00 to 1000 grams per metre.
~. .
: .
' W092/02301 PCT/G~I/Ot293 ~,~

Of course, any of the wires outlined above can be used together by, for example, being knitted togeth~r or b~ing incorporated into different gauzes installed contiguously to form a tailored pack (either with or without a separa-ting 5 inert gauæe, felt pad or knitted screen between layers of the pack)~ Also, knitted ~auzes formed from any of thQ aboYe ~.7ires can be encapsulaked by a catalytic mesh to form a cartridge that looks essentially similar to e~lsting cartridges. Knitt~d gauzes can also be supported by an inert screen.

It is envisaged that knitting will be especially sui~able ro, the production of catchment gauzes. Whilst any of the stitches disclosed hersin may be used for catchment g~u2~s, jacquard knits in particular would be advantagPous in Yiew of t~ei-'three-dimensionality'. In other words, each layer of jacquard 15 knitted material has a marked thickness which means that post-reaction gases ake longer to pass through the layer, thereby increasing its catchment performance. Jacquard knitted ~ :
material may al50 improve the cataly~ic performance of catalytic layers, for the same reason.

20 The limitations of cixcular or rotary knitting machines in terms of lac~ of width tend to be less acut2 in the fi~ld oL
catchment gauzes. This is becaus~, in soma applications catchment gauzes are hidden out of sight within a pacX and therefore the presence of seams is relatively unimportant.
25 Indeed, it is possible for the tubular knitted product simply to be folded over and packed down to form a catchment layer.
The advantages of circular or rotary knitting machin~s, such as speed and flexibility, can therefore be enjoyed under such circumstances without significan~ penalty.
':
30 A further potential benefit of knitting s~ems from the fact that the characteristics of the wire (diameter, composition and so on) can be xeadily varied during knitting simply be substituting a new wire. This facilitates the production of tailored packs, in which the characteristics of the wire are .'.: ,, ~ ' . . , '. : . ' ~ ' ' ' t`2ilorad to suit the conditions (of ma~erial loss etc.) prevailing at different points through ~he pack. Thus, the layers of the tailored pack can be knitted in a continuous operation, even though each layer requires a differ~nt wire.

5 Another advantage of knitting is that catalyst packs can be produced to a desired thickness without havir. to form layers of gauzes or felts. This helps to reduce production costs and ~lso minimises expensiYe installation time. Th~ knitting process can be similar to that used in the manufacture of lO lace, in which additional material is knitted on to a knitted or woven backing to produce r~gions of a desired thickness.
Another way cf thickening a layer is to produce a Ipile~
similar to velvPt or velour cloths.

A knitted layer may have non-uniform thicknQss so that, for 15 example, the central region of a layer is thicker than the ou~er region of that layer. Such an arrangement may be advantageous as it allows the residence time of ~ases passing through the layer to be tailored to suit a velocity gradient across the width of a reactor. For example, the velocity of 20 the gas stream is generally lower adjacent the walls of a reactor than towards its centre. Hence, the centre of a layer may be thickened so that the residenca time i.s more nearly uniform across th~ layer. It is similarly possible to vary the thicknesses of the individual wires that make up the layer, 25 so that the catalytic effect varies across the layer.

A further possi~ility is to thicken the edge or edges of a layer to compensate for the edge erosion that is ~ometimes observeA. At present, reinforcing patches are commonly used to res e and/or tc reinforce the edges of a catalyst pack.

30 Of course, arrangements combining a thickened edge region and a thickened central region are also possible, as are arrangements combining a variable-thickness layer with variations in the thickness of the wixes making up the layer.

WO92102301 ~ PCT/GB91/012~3 ~ 12 -Embodiments and a5pects of this invention will now be descrlbed, ~ y of e~ample only, with reFerence -~o ~he accompanying drawings in ~hich:

Fig. 1. shows a partial cross-sectional view of a prior art 5 uoven gau~e useFul in a~oni~ o~dation c2tal~st or getter systems;

Fig.2. s:~cT~s a ?~ia_ c-oss~secs cnal ~Jie~- of a single fac2 knitted te~tile according to -che in~nt1on, made on a sin~le- :
bar wa~p 'cni~iny machine;

10 Fig.3. shows ~ 2r. ial cross-s~c ional vle.-; of a dou~le face knitted tex il_ acco-ding lo ~he invention, made on a warp knitting machine;

Fig. 4. shows a knitted textile being made on a warp knitting machine with bearded needles that stitch in unison; -:
15 Fig. 5. shows a partial cross-sectional ~ie~ of a preferred kni~ted textile having a tricot stitch, as made on a Mayer warp knitting machine; and Fig.6. shows a partial cross-sec~ional view of a netting type stitch made on a two-bar Raschel knitting machine.

20 Knitted precious metal textiles according to the invention can ~.
be made from single or multiple strands of wire, each strand ~ :
having a diameter of about 0.05 mm to 0.10 mm. Representative : ::
knitted textiles of the invention include those with wires having 2 diameter of 0.06 mm, 0.675 mm, 0.7~ mm, 0.09 mm, and 25 combinations thereoI. The wire can be made of any precious mPtal ~this term including precious metal alloy). Particularly suitable catalyst results have been achi~ved with platinum wire, especially platinum alloys that contain more than 50 ..
perce~ t platinum. Representative alloys with rhodium and : .~
.,,; :.....
..

WO 92/02301 ~ if~ PCI`/~B91/01293 palladium are shown in Table 1.

T~BT,E I

Com~osition of Wires Used in Rnitted Metal Textiles ~ ~atinum Rhodium Palla~ium 5 1 95s ~ô
2 90~s ~S ;%

3 90~ îO~

4 92%
9~2Q;, 3 J 5 4 ~) 5 10 6 gO-9S~ 2.~'s 2.
7 97~ 3 8 75~ ~ 21~

~hen a t~xtile for use in a getter sy~tem ic de~ired, it has been found that wires made of palladium and palladium alloys 15 axe preferred, especially alloys that contai~ more ~han 50%
palladium.

Certain knitti~g patterns are especially advantageous, including jacquard, raschel an~ tri~ot designs. These patterns, which contain a large number of large knots 20 surrounding large holes, permit a large quantity of wire to be u~ed in a small space, without closing up the holes and causing bacX pressure problems. Thus, in comparison with ~ conventional woven gauzes, more catalyst or getter material can be used in the same space. This results in improved 25 reaction efficiency, without blocking the flow of reaction .
: gases.

l'he structure of a conventional woven gauze is shown in Fig.l. ~:
These gauzes are e~tremely simple, and are made o~ overlapping ;: perpendicular strands (a,b) which lay next to each other, but 30 which axe not looped, stitchQd, !cnotted, or otherwise bound ~ .

WO 92/02301 , ~ PCT/GBgl/01293 - 14 ~
~o each other. Noven material is quite different even from a relatively simple single face knitted fabric made on a single bar warp knitting machine, as shown for e~ample in ~ig 2. A
double face warp knit (having loops in two directions) further illustrates the differences in structure and complexity between prior art woven gau~es and the knitted precious metal textiles of the invention, as shown in Fig.3. Indeed, si~en the relatively fragile nature of precious metal ~ires and their high costs, it was thought prior to this invention tha~
10 knitted precious metal catalysts and getter systems could no~
be suitably made. It was believed that the complexity of the knit compared to weaving would result in brea~age, une~
results, and difficult problems of control that would rende~
knitting impractical, and mors costly and time consuming than 15 weaving. However, it has now been discovered that precious metal wires can be knitted according to the invention .into textiles that are superior in strength and construction to conventional woven gauzes. Furthermore, it has been discovered that knitt~d textiles can be mad~ much f aster than woven 2 0 gauzes, without breakas~e and quali~y control problems .

According to the invention, any conventional knitting machine can be adapted to produce the textiles of the invention, including straight knitting machines, circular or rotary knitting machines and warp knitting machines. Multiple needle 25 machines are preferred, so that large textiles can be rapidly made, and warp knitting machines have been found to produce especially ad~antageous results. Fig. 4 illustrates a knitted precious metal textile being made according to the invention on a ~ayer warp knitting machine with latched bearded needles.
,. ..
30 In addition to separate catalyst and getter textiles, the invention also contemplates "self-get~ering" catalyst textiles, in which a catalyst material and a gettering material are knitted into a single composite textile. ~or example, catalyst strands comprising platinum as the 35 predominant material can be knitted with getter strands ' .
;'.

WO92/02301 2 ~ 8 81.~ ~ pcr/GB9l/ol293 ~ 15 -comprising palladium as the predominant material, to pro~ide an integral textile material that both catalyses ammonia oxidation reactions and iYnmediately recaptures volatile catalyst material (such as platinum) which might otherwise be S lost. Thus, one embodiment of the invention contemplates a new material that would permit a single textlle to serve two purposes. With such a textile, conventionally separate downstream gettQr gauzes would no longer be needed. According ~o an aspeot of the invention, metal losses can be rPduced at 10 or very near the source of loss by knitting the ca~alyst and entrapment systems into one textile. As just one example of how this can be done, a single continuolls textile can be knitted, which has ~lternating sections of predetermined length of catalyst and getter material, the final textile car 15 then be folded into a pac~, comprising adjacent alternating layers of catalyst and getter ~s desired~ The resulting self~
gettering cataly6t can then be conveniently supplied as a unitary cartridye.

The invention is also suitable for knitting ~ailored catalyst 20 packs, which pro~ide redu~ed weight of the s~me efficiency by using tailored overlapping shapes.

In another embodiment, a textile can be knitted using strands of different thickness along the length of the material ~ a resul~ which cannot be achieved by conventional weaving, since 25 thread sizes on a loom cannot be readily changed anywhere in the textile as they can in a knitting process. Thus one advantageous product according to the invention, uses thinner wires at the top and thicker wires at the bottom, to provide a knitted ! ecious metal textile with a ~trong pyramid-type 30 infxastruc~ re. This thickness pro~ile can also be reversed or otherwise varied to produce a pack whose wir~ thickness at a given point in the pack is tailored to suit ~he metal loss rate expected to that point in ~he pack.

To make the knitted precious metal textiles of ~he invention, WO92/02301 ~ PCT/GB91/01293 ~9 16 -it has been necessary to overcome cerkain problems not encountered in conventional weaving processes, and n~t encounterod in traditionzl ~a~ic Xnitting processes. Unlike weaving, which does not loop strands of wire and does no~ use

5 knots and bin~ing points, the knitting ~rocess draws strands of wlre into sharp loops, through which needles carrying other strands are passed, to form a looped stitch. I~itially, attempts to knit precious metals ~ere unsuccessful, due to constant b,3a~age o' -~he wi.e 'h-~ad and j~ming and selzing lO problems ~ h ~'ne ~2ri ~US !~r.i_~ing m2chines tha~ were tried.
The resulting textll~o ~roduc~s~ere une~Jen and unsatisfactory.
After many 2~pe~ ents, ~nd T~i~h m~ch ~rial and error, it was discovered that knitted precious metal textiles can be made using prec~ s meta1 a'lo-,~a of ~la ~n~m,/ rhodium and/or 15 palladium, in ~hic'~l._sses ,_ng~ng -rom 0.~5 -to O~lO mm. This can be achieved either by lubricating the wire with a lubricant that does not interfere with the knitting machine, or by using a transitory feed thread to guide the wires. Where a transitory feed thread is u~ed, it is advantageous for the 20 thread ~o be of copper alloy: this can be etched away a~er knitting.
' :.
In some cases, it is also necessary to opera~e the knitting machine at slower speeds than are commonly used for Xnitting fabrics or non-precious metals. Suitable lubrican~s include 25 spray starch and spray wax. Even at slower speeds, a lubricated metal textile can be knitted as much as ten times faster than conventional catalyst gauze~ can be woven on a loom.

Knitting machines using independently moving needles, or 30 latched needles which move in unison can be used, and straight, circular, or warp kni~ting machines are suitable, though warp knitting machines are preferred. Xnitted metal textiles of the i~vention can also be made ~ith loops which follow either the length or width of the fabric. In warp 35 knitting machines, a large quantity of parallel threads ,.' .
': .

WO92/02301 ~ ~ .)0 PCT/GB91/01293 running lengthwise are fed to one or more needle bars, with collectively moving needles. These threads are laid around the needles by loopovers to sim1~ltaneously form stitches across the width of the material. Si~gle ~ace and dou~le face knitted fabric can be used, although single face embodimen~s such as tricot are easier to produce.

The invention is further described wi~h re~erence to a number of e~amples. It is LO ~ understood tha L th2se examples are illustratl~, anà do not li~i-L th_ scope of the ~ppended lO claims.

Example l A Xnitted t2~ctile a~cordi~g ~o the invcntion ~as made on a Mayer warp knitting machine, which permits a standard knitting matrix of about 40 x 40 wires. See, for example, Figure 4 of 15 the drawings. The machine can be modified, however, to produce other matrices as desired. Depending on the chosen stitch, a square, oblong, or other lattice can be made. Mayer machines most suitabla for knitted catalyst and getter textiles are available to knit in widths of 84 inches and 210 inches, at 20 speeds of about 50 to lO0 feet per hour. The preferred textile made on the Mayer machin~, according to ~he invention, is a tricot pattern, as shown in Fig. 5 .

In one embodiment fo'und to be suitable, wire having a diameter of .003 inches ~0.076 mm) and a composition of 90% Pt 5~ Rh -25 5~ Pd was knitted on an 84 inch machine into a textile 36inches lon~. An acceptable product wi~h a ~ine uniform weave was achieved by lubricating the wire with spray ~tarch, and by feeding the wire through tension controllers in conjunction with the actual knitting. The best results ~ere achieved using 30 double bar operation, which produced a strong, durable and flexible knit, although single oa' operation is somewhat faster and easier to manage.

.:

WO92/02301~ PCT/GB91/0i293 - One sample of warp knit tricot material ll ~ 13 inches long, weighed 53.6 g and had a density of 583 g/m2, which is suitable for catalyst and getter applications.

Example 2 5 As shown in Fig.6, a Raschel knitting machine can be used to ~ni~ a netting-type pattern, using two-bar cons~ruc-~ion.

E~am~le_3 A knitted precious metal textile was made on a Tritex circula,.
or rotary knitting machine, which produces 2 closed tu~e or 10 sock of material having a netting-liXe appearanc~. The Tri tex machine produces a tube with a maximum diametex ~f 30 i~ches, which can be slit to form a textile produc~ that is either 94 inches in one layer, or 47 inches when doubled. The Tritex machine produ~es a maximum knitting matrix of 25 x 30 wires, 15 which results in a relatively open knit in the final product.
The Tritex textile was made of wires having a diameter of 0.003 inches, and a compositiun of 10% Rh-Pt. Using a 30 inch machinet khe output is about 33 feet per hour.

Example 4 20 A knitted metal textile was made on a Lamb circular knitting machine. In this embodiment, using a 10 .inch machine, it was necessary to run the precious metal wire with a polyestex lead thread, to avoid snagging and breakage of the wire during loop formation. Multiple strands of wire can be knitted 25 simultaneously in this way. Other lead threads, such as nylon, ~: cotton, xayon or the like, can also be used. The lead ~hread can be dissolved or burned away prior to use, or during final ~ flame activation of the material r or even with the first on-:~. site use. A tubular sample of this textile that was 1~ inches 30 wide and 15~d inches long weighed 14.5 g, and had a density of 418 g/mZ. The density can be altered by increasing or ., .' .":

WO 92fO23~1 2 Q 3 ~ P~T/GB91/01293 decreasing the number of needles, or by adding ox subtracting strands of wire.
.

' .:

, ~ : : . . .

Claims

1. A knitted textile product made of precious metal wires.

2. A knitted textile product according to claim 1, comprising jacguard, raschel or tricot stitches, or any combination thereof.

3. A knitted textile product according to claim 1 or claim 2, whose thickness is non-uniform.

4. A knitted textile product according to claim 3, whose thickness increases moving inwardly away from its edges.

5. A knitted textile product according to claim 3 or claim 4, whose edge regions are thickened.

6. A knitted textile product according to any preceding claim-, having non parallel side edges when uncut.

7. A knitted textile product according to any preceding claim, whose wire density is non-uniform.

8. A knitted textile product according to any preceding claim, whose wire characteristics are non-uniform.

9. A knitted textile product according to claim 8, wherein the differing characteristics include wire diameter or wire composition.

10. A knitted textile product according to any preceding claim, comprising catalyst wires and getter wires.

11. A knitted textile product according to claim 10, wherein the catalyst wires are predominantly platinum and the getter wires are predominantly palladium.

12. A knitted textile product according to any preceding claim, wherein the precious metal wires are of intrinsically catalytic material.

13. The use of a knitted textile product defined in any preceding claim, as a catalyst or as a getter, or as a self-gettering catalyst.

14. Any substance produced by the use of a knitted textile product defined in any of claims 1 to 12, as a catalyst or as a getter, or as self-gettering catalyst.

15. The use of a knitted textile product defined in any of claims 1 to 12, in the synthesis of hydrogen cyanide from ammonia and methane.

16. Hydrogen cyanide synthesised from ammonia and methane by use of a knitted textile product as defined in any of claims 1 to 12.

17. A catalyst pack containing a plurality of layers, wherein at least one of the layers is a knitted textile product according to any of claims 1 to 12.

18. A catalyst pack according to claim 17, wherein the layers are formed by a continuous knitting operation.

19. A catalyst pack according to claim 18, wherein the layers are produced by folding a single knitted textile product.

20. A catalyst pack according to any of claims 17 to 19, wherein the wire characteristics vary from one layer to another.

21. A catalyst pack containing at least one concealed catalyst or catchment layer, the concealed catalyst or catchment layer comprising a precious metal knitted textile product produced on a rotary knitting machine.

22. A catalyst pack according to claim 21, wherein the concealed catalyst or catchment layer comprises a plurality of pieces of rotary knitted product joined together.

23. A catalyst pack according to claim 21 or claim 22, wherein the concealed catalyst or catchment layer is formed by folding over and packing down the knitted textile product.

24. A method of manufacturing a precious metal textile, comprising knitting together precious metal wires in a knitting machine.

25. A method according to claim 24, comprising lubricating the precious metal wires before and/or during the knitting process.

26. A method according to claim 24 or claim 25, wherein the knitting machine is a warp knitting machine.

27. A method according to any of claims 24 to 26, comprising varying the width of the knitted product emerging from the knitting machine, to produce an uncut knitted product having non-parallel side edges, such as a circle.

28. A method according to any of claims 24 to 27, comprising varying the thickness of the knitted product to produce a knitted product whose thickness is non-uniform.

29. A method according to any of claims 24 to 28, comprising varying the characteristics of the wires during the knitting process.

30. A method according to any of claims 24 to 29, comprising varying the wire density of the knitted product by varying the number of wires fed to the knitting machine.

31. A method of manufacturing a layered catalyst pack, comprising knitting the layers of the pack in a continuous operation and then folding the knitted product to form the layered pack.

32. A method according to claim 31, comprising varying the characteristics of the wires from one layer to another, thereby to produce a pack whose layers have different characteristics.