CA1288561C - Aldehyde cross-linked porous membranes - Google Patents

Abstract

ABSTRACT
A polymeric porous membrane having a matrix made from an aliphatic thermoplastic polyamide or from an aliphatic thermoplastic polyamide/polyimide copolymer which has both relatively non-crystalline and relatively crystalline portions is provided. A method for preparing such a membrane comprises dissolution of an aliphatic thermoplastic polyamide or polyamide/polyimide copolymer having crystalline and non-crystalline portions such that the non-crystalline portions dissolve. At least some of the crystalline portions form a colloidal dispersion which is formed into a film. Dissolved non-crystalline portions are subsequently precipitated to form the porous membrane matrix. The pores in the membrane are defined by spaces between the relatively crystalline portions and at least some of the relatively crystalline portions are linked together by the reaction of a bis-aldehyde with the membrane matrix.

Description

1'28BS~

This invention relates to porous membranes made from aliphatic thermo-plastic polyamides or aliphatic polyamide/polyimide copolymers.

Synthetic polymeric membrcmes are used for separation of species by dialysis, electroclialysis, ultrafiltration, cross flow filtration, raverse osmosi~ and other similar techniques. One such synthetic polymeric membrane is disclosed in Australian Patent Specification No. 505,494 of Unisearch Limited.
The membrane forming technique disclosed in the abovementioned Unisearch pat~nt specification is broadly described as being the controlled uni-directional coagulation of the polymeric material from a solution which is coated onto a suitable inert surface. The first step in the process is the preparation of a "dope" by dissolution of a polymer. According to the specification, this is said to be achieved by using a solvent to cut the hydrogen bonds which link the moleculax chains of the polymer together.
After a period of maturation, the dope is then cast onto a glass plate and coagulated by immersion in a coagulation bath which is capable of diluting the solvent and annealing the depolymerised polymer which has been used. According to the one example given in this specification, the "dope"
consisted of a polyamide dissolved in a solvent which comprised hydrochloric acid and ethanol.
In another membrane forming technigue, the liquid material out of which the membrane is cast is a colloidal suspension which gives a surface pore density that is significantly increased over the surface pore density of prior membranes.

..

.
-5i6:3 According to that technique a thermoplastic material ~aving both relatively non-crystalline and relatively crystalline portions is dissolved in a suitable solvent under conditions of temperature and tirne which cause the relatively non-crystalline portions of the thermoplastic material to dissolve whilst at least a portion of the relatively crystalline portion does not dissolve but forms a colloidal dispersion in the solvent. The collQidal dispersion and solvent (i.e. the "dope") is then coated onto a surface as a film and thereafter precipitation of the dissolved thermoplastic portion is effected to form a porous membrane.
Membranes of both of the above kinds suffer from disadvantages which limit their commercial usefulness and applicability. For example, they exhibit dimensional instability when drying and may shrink by up to 7~. Thus, it is essential that they be kept moist prior to and after use. Furthermore, where the membranes are made from polyamide, it has not been possible to generate concentrated and varied chemical derivatives of the membranes and this restricts the situations to which the membrane may be applied.
Another disadvantage is that such polyamide membranes are fundamentally unstable and eventually become brittle on storage. The instability has been carefully investigated by I.R. Susantor of the Faculty of Science, Universitas Andalas, Padang, Indonesia with his colleague Bjulia. Their investigations were reported at the nSecond A~S~EoA~N~ Food Waste Project Conference", Bangkok, Thailand (1982) and 30 included the following comments regarding brittlenesso "To anneal a membrane, the thus prepared membrane ~according to Australian Patent NoO 505,494 using Nylon 6 yarn) is immersed in water at a given temperature, known as the annealing temperature, T in degrees Kelvin. It is allowed to stay in the water a certain length of time, called the annealing time.
For a glven annealing temperature, there is a maximum .
: ~ :
. .

~8~6~L

annealing time, t(b) in minutes, beyond which further annealing makes the membrane brittle. Plotting ln l/t(b) versus l/T gives a straight line. From the slope of this line it can be concluded that becoming brittle on prolonged annealing is a process requiring an activation energy of approximately 10.4 kilocalories/mole. From the magnitude of this activation energy, which is of the order of van der Waals forces, the various polymer fragments are probably held together b~y rather strong van der Waals forcPs or hydrogen bond(s)."

We have confirmed that the brittleness is due to a recrystallization of water-solvated amorphous polyamide.
In some cases tsuch as polyamide 6) brittleness occurs within 48 hours of immersion in distilled water (p~7) at 80C. Colorimetric ~NH2 end group analysis has shown that there is no significant hydrolysis of the amide groups during this time. As would be expected, the rate of embrittlement is catalysed by dilute acids (eg: pH of 1.0) due to nitrogen protonation and subsequent solvation. This ~ effect explains the apparently low "acid resistance" of the -~ polyamide membranes. However colorimetric determination o both -~2 end groups and -COOH end groups has shown that the effect is due to crystallization rather than acid ~5 catalysed hydrolysis.
~ Th~re is a potential source of confusion in the use of words such as "acid-resistance~ in the context of this specification. That most of the brittleness is due to physical effects rather than chemical decomposition or chemical solvation (at least for dilute acids) is shown by - the extreme embrittlement caused on standing 5 minutes in absolute ethanol. The ethanol removes the plasticizing water tenaciously held by non-crystalline nylon as will hereinafter be described in relation to Example 2.

.
:,.

, s~

Acc~rdingly, the following definitions apply in this specification:-(a) "Embrittlement resi~stance" means hindrance or prevention of the physical recrystallization mechanism of the amorphous polymer matrix~

(b) "~cid-catalysed embrittlement resistance"
means prevention of embrittlement of type (a) even in the presence of dilute acids (pH 1 to 7).

(c) "Acid solubility" means the rapid dissolution of polyamide in strong acids (100~ formic acid or 6 N hydrochloric acid).

(d) "A~id catalysed hydrolysis" means the scission of amide bonds (such scission is much faster in l; an amorphous polyamide than in a crystalline polyamide.) As well as "embrittlement" the prior art me~branes show the normal chemical defects of the parent nylon polyamides in that they possess only modera~e oxidation resistance and bio-resistance.
It is an object of this invPntion to provide Rolymeric porous membranes composed of thermoplastic aliphatic polyamides ~including polyamide/polyimide copolymers) which have greater resistance properties and 2~ improved mechanical stabilîty than prior art membranes. It is a further object of the invention to provide polymeric porous membranes which readily lend themselves to the preparation of chemical derivatives thereof for particular uses.

.. .
; .. . . -, ~ , . . - , :. . . :, : : , :

~2~3~35~i~

~ ccordlng to one aspe~t of the invention there i~
provided a method o~ prepar ing a porc~ membrane compr is ing the ~tep3 o~ :-1 i ) di~solving an aliphatic polyamide or an aliphatic polyamide/polyimide copolymer which h~ both relatively non-crystallin0 and relatively crystalline portion8 into a ~olvent under conditions of temperature and time which cau~ the r~latively non-crystalline portion~
of thls polyamide or copoly~r to di3~01~e while . at le~st a part of the r~lat~v~ly crystalline portion~ o~ th~ polyamide or ~opolymer do ~IOt di3~01v~, but form a colloidal di~persion in aid ~olvent, ( i i ) forming said ~olloldal di~per~ion and ~olvsnt into z~ filan ~i~d thereafter causing pr~cipitation of at least part of the di ~olved non-cry~t~lline porl:iolls in the ~ilm to form a poro~3 r~embralle in which th~s pores are def ined by ~pace~ b~tween the relatively c:rystalline portiona~ nd~
( iii ) reacting the membr2~ne with ~ polyfunctional aldehyde ~ as herein defined to llnk ~t least some of the relatively crystalline portions with the aldehyde.

In the context of this spscificatlon the term "a polyunctional aldahyd~ as herein de~ined" means an aldehyde or aldehydc yieldln~ ~i~ture in which the ald~hyd~ functionality exceeds one -- C~ ~ O per molecule.
Preerably, the polyfunctional aldehyde is a bis-aldehyde and is selscl:ed from the group comprising glutaraldehyde, glyoxal, succinic dlaldehyde, alpha-hydroxyadipic aldehyde, terephthalic dialdehyde and phthalic dialdehyde as well as ~ ., .

. .

t .
:, , .
,' s~

mixtures thereof. Furthermore, the aldehyde may be derived from a bis-aldehyde polymer, an acetal or an acetal ester.
The aldehyde reaction step may be controlled so that from 10% to 25~ of the aldehyde chains are not linked at one end. In which case, the invention can include any one of the following steps of reacting at least some of the free ends of the single-link aldehyde chains with:-(i~ a phenol that may be selected from the group comprising resorcinol, diphenylolpropane, tannic acid, pyrogallol, hydroquinone, meta-cresol and naphthol as well as derivatives or mixtures thexeof~
(ii) a protein such as gelatin, (iii~ a polyhydric colloid such as hydroxyethylcellulose, or (iv) an amine such as melamine.

The phenol modified chains may be further reacted with:-(a3 sodium monochloroacetate in agueous solution, or (b) aqueous diazonium salts Also, the phenol modified chains may be subjected to further processing including the steps of:-(a~ reacting at least some of the remaining reactive single-link aldehyde chains with hydra~ine, (b) reacting the phenolic hydroxyl groups with epichlorohydrin, (c) reacting the resultant epoxides with a diamine to fix a pre-determined concentration of amine groups, hydrolysing excess epoxide groups to hydroxyls and r - :' ' ' , ' :

-. . ~ - .

(d) reacting the amine groups with excess bis (isothiocyanate).
A membrane made in accordance with the method of the invention may be further treatecl by reacting it with sodium bisulphite, hydroxylamine-O-sulphonic acid or phenylhydrazinesulphonic acid. A phenol-modified membrane may be further reacted with a bis-aldehyde.
The invention also provides a polymeric porous membrane comprising a membrane matrix made from an aliphatic thermoplastic polyamicle or from an aliphatic thermoplastic polyamide/polyimicle copolymer which has both relatively non-crystalline and relatively crystalline portions and in which the relatively crystalline portions are joined together by relatively non-crystalline portions with pores in the membrane being defined by spaces between the relatively crystalline portions characterised in that at least some of the relatively crystalline portions are linked together by the reaction cf an aldehyde as herein defined with the porous membrane matrix.
A particularly preferred bis-aldehyde is the five carbon atom glutaraldehyde which has the following formula:-C) C--CE~2--- CH2--CH2-- C-- ' ' ' H H
~ When polyamide 6 is used as the polymeric membrane, each low or relatively crystalline chain has a number of amide groups spaced apart alor,g the chain and the bis- :
aldehyde (such as glutaraldehyde) displaces the hydrogen atom of the amide groups with their end carbon atom becoming bonded to the nitrogen atom in the polyamide chain as follows:-N -~CH - CH2 ~ CH2 CH2 - CH - N

:

:~L2~

The glutaraldehyde provides a true cross link between the polyamide chains and this increases the membrane's bio-resistance as well as its embrittlement resistance.
In a modification of the invention, from 10 to 25% of the glutaraldehyde chains are not linked at each end to a polyamide chain but rather one end is unattached to leave the CH = o group in a more react:ive form. This mcdifi~ation further improves the dimensional stability of the membrane and allows extensive chemical modification.
Steric effects ensure that there are always some detectable singly linked glutaraldehyde molecules but the proportion can be increased by using conditions which favour rapid reaction and by high glutaraldehyde concentration~.
A feature of the glutaraldehyde type of cross-linking is that the permeability of the original polyamide membrane to water is unexpectedly only ~lightly and controllably affected as will be hereinafter apparent from example 2, although (as expected) the permeability of many dissolved solutes is greatly affected.
Further reaction with a phenol provides a membrane having acid-catalyzed embrittlement resistanceO The resultant polyamide/phenol~aldehyde block copolymer is particularly useful in the treatment of effluent from food processing plants where alkaline mixtures are used as a cleaning agent, often after an acidic enzymatic cleaning treatment.
When the free end of the single-link aldehyde chain is reacted with a phenol (such as resorcinol) the free end of the aldehyde chain is transformed to bis (phenylol) methane:-QH
~ OH
- CH=O ~ CH ~
~ OH
OH
In addition to glutaraldehyde (or other bis-aldehyde) a small amoun~ of formaldehyde may be used as the link particularly if free ends are reacted with resorcinol.
All or part of the glutaraldehyde can be replaced by ,: . : . . : . .
-' . : ' . , - . ~ :
.
,, : :
: . :: - ' '. . -: , :

~8563~
g equivalent amounts of many commercially available bisaldehydes 3uch as glyoxal, ~uccinlc dialdehyde, alpha-hydroxyadipic aldehyde, terephthalic dialdehyde and phthalic dialdehyde with very ~imilar results including the preparation of chemical derivatives of the membrane arising from a proportion of end groups reacting as an aldehyde. Choice of bis-aldehyde dep~nd~ largely on economic, ~afs-handlinq and aldehyde storage stability factors rathex than chemical reactivity for mo~t applications. Nevertheles~ ~ome quite ~ubtl~ differences ~uch as ab~orp~ion of colloid~ which can be important in commercial u~age of the membrane may affect the choice of aldehyde.
The aromatic aldehydes are ~lower reacting, giving lighter-coloured products and are harder but ~ore lS brittle. They also ~how the usual difference~ ~hat aromatic aldehyde~ ~how from aliphat.ic aldeh~des, eg:
slower.reaction with bi3ulphite.
Any de~ired properties likely to be needed in ultrafilters, ion-e~change re9ins, ion-specific resins, dyeing colour (by reaction with diazonium salts) or intermediates for highly active enzyme immobilization or affinity chromatographiG ~urfaces can be obtained by choosing a cheap glyoxal, glutaraldehyde, succinic dialdehyde or terephthalic dialdehyde and combining with a cheap reactive 2~ phenol ~uch a~ r~oxcinol, diphenylolpropan~, hydr~guinone, pyrogallol, tanni~ acid or naphthol as well as mixtures or derivatives thereof. For speci$ic purpo~e~, specific phenolic derivatives can be u~ed or the pra erred glutar~ldehyde/re~orcinol treatment can be modified by simple ~oaking proceduxe~ in appropriate reagent~.
A ~table, ~terilizable, controll~bly porou3 structur~ can b~ made by sequential reaction a~ in Example 6 with hydrazine, epichlorohydrin, h~xamethylene diamine and 1,4 - phenylenebisi~othiocyanate. This is excellent for reaction with the -N~2 end grcup~ of many proteins, whil~t still ~llowing bioactivity and aff inity , ,~, ., `.

:

chromatography for harvesting anti-bodies. The protein bond is covalent and stable but on a suitably long arm on an extended controllable interior structure.
In contrast thereto, a glutaraldehyde treated polyamide may be made electrically conductive by treatment with 4-phenylhydrazine-sulphonic acid to provide an electrodialysis membrane when the porosity is almost zero to a hydraulic pressure difference as in Example 3.
The reactivity of products containing highly reactive aldehyde groups is not restricted to phenols, although the latter are preferred for long-lived and aggressive environments. For example, prot:ein, gelatin or hydxoxyethylcellulose can be reacted with the membrane to give products which are very elastic and rubbery in ethanol. Furthermore, the preferred properties of the glutaraldehyde/resorcinol treated membranes can be combined with free aldehyde group reaction versatility by reacting once again with a bis-aldehyde to give an enhanced free aldehyde content. The product is then a polyamide/glutaraldehyde/resorcinol/bis-aldehyde which can form more concentrated and more stable derivatives. There is some advantage in using glyoxal for the last bis-aldehyde to give highest concentrations of -CHO. However, glutaraldehyde seems best for initial reaction with the polyamide, presumably for steric reasons of cross-linking.
Of course it is possible to involve the use of small quantities of the cheap mono aldehyde, formaldehyde, at various stages to dilute the bis-aldehydes. However for steric reasons formaldehyde is undesirable for cross-linking in the initial polyamide reaction. Formaldehydecan have some use for a -further diluent reaction with phenols. However, it is preferable to condense the ~ormaldehyde separately with the phenols to make controlled pure reagents or condensation products and then to condense these with the polyamide/glutaraldehyde precursor.

''' '- ' . ' ' '' . ~ -:
, ' ' ' . ' : - :

EX~MPLE 1 A solvent (A) was prepared by mixing 225ml of 6.67N
hydrochloric acid with 15ml of anhydrous ethanol. 90 grams of 55 dtex 17 filament polyamide 6 with ~ero twist (which constitutes the polyamide starting material) was added to solvent A held at a temperature of 22C over a period of less than 15 minutes.
The dope of the polyamide 6 and solvent A was then left to mature for 24 hours at a temperature of 22C during which the relatively non-crystalline portions of the polyamide dissolved as did no more than 50% of the relatively crystalline portions of the polyamide 6 with the remaining relatively crystalline portion dispersing in the solvent.
After maturation, the dope was then spread as a film of about 120 micron thick on a clean glass plate. The coated plate was placed in a water bath where precipitation of the dissolved portions of the polyamide was effected ; 20 within 3 minutes. The membrane was then reacted with 5~ of glutaraldehyde tbased on the dry membrane weight~ at a pH
of 3 to 6 at a temperature of 60C overnight. It was found that 50~ to 80~ of the glutaraldehyde had reacted depending upon the pH and that of these percentages 10% to 25% of the glutaraldehyde had one aldehyde free for further reaction.

.
The polyamide 6 membrane made according to Example 1 had a water permeability of 339 litres/square metre/hour and rejected 81% of the protein in a standard edible gelatin.
60 grams o this membrane were treated with 2.24 grams of glutaraldehyde in 138 grams of water at p~ 5.5 and at a temperature of 20C for 1 week followed by water washing.
It was found that the membrane had reacted with 2.7% of its dry weight of glutaraldehyde. Of this 2.7%, about 0.6~%

:

.. . . . ..
- :

.. : . :
.. - .. . : - .. , . ~ :.. . . :
- , : ...... . .

5~

(ie: 23% of glutaraldehyde reacted) was still reactive as an aldehyde. The water permeability was now 384 litres/square metre/hour and the gelatin rejection was 82~. These differences from the original permeability and gelatin rejection figures are very slight ~or any practical use.
The cross-linked membrane was not "acid-catalysed embrittlement resistant" as it became brittle in 6 days at 60C at pH l but was unaffect2d in 35 days at 60c at pH 13 (alkaline). There was a complete absence of traces o~
terminal -NH2 groups, originally present in the polyamide 6 membrane of Example l, as shown by the disappearance of the original yellow reaction with D.A.~.I.T.C. reagent, 4-dimethyl-aminophenylazobenzene - isothiocyanate. The slowed "acid-catalysed embrittlement" is due to the slow reversible reactions which yield glutaraldehyde and the starting polyamide 6. Such reactions are due to the acid~
labile group, ,, I
N - CH ~ NH + O _ CH

OH

Also there will be present some proportion of acid-~abile glutaraldehyde polymers. Confirmation of this reversibility was shown by the reaction of the membrane of this Example 2 with M/400 2,4-dinitrophenylhydrazine tDNP~
in N/100 HCl at 22C. In 21 hour 15% o the total glutaraldehyde had reacted with and removed from solution an equivalent of DNP; in 37 hour 17~4% and after 48 hour at 60C, 23%. The reaction of glutaraldehyde with primary amides -CO -NH2 has been well studied and the products are said to be stable reactive gels for affinity chromatography;see P. Monsan, G. Puzo and H. Mazarguil, Biochemie, 57, pl281 tl975). Reaction of polyamides containing secondary amide -CO-MH with glutaraldehyde could be expected to give less stable products.

~
: : .

s~

Despite the low "acid catalysed embrittlement resistance" of the abovP polyamide 6/glutaraldehyde reaction product it was found to be a key intermediate in the preparation of preferred stable, tough, rubbery ultrafilter membranes by reaction with resorcinol (see Example 5) and of stable tough, rubbery oil and detergent repelling ultrafilters by reaction with gelatin or hydroxyethylcellulose (see Example 7). The reaction with gelatin in Example 7 illustrates the method of immobilizing an enzyme and for preparing absorbents for affinity chromatography. Many useful chemical derivatives can be prepared by known procedurPs and are described in examples below.
The brittleness of the membranes of examples one and two air dry ~70% Relative Humidity) wet and in ethanol are indicated by the extension to break on stretching and by behaviour on rubbing in the following tables:-MEMBRANE - EXAMPLE 1.
_ , _ . . _ Air Dry ~et Ethanol . .. _ . __ -Extension to Break 10% 60~ ~0 _ _ _ _ Behaviour on rubbing Rubbery Rubbery Powdered .: : .. . . : .
; : - ~ -- . : , : . ' .- : , ' ~ ' ' :. ' . : ' ~8S~

Air Dry Wet Ethanol Extenslon .
to Break 6~ 60% 20 . _ _ Behaviour on rubbiny Rubbery Rubbery Powdered Thus, the glutaraldehyde to this stage has altered the chemical rather than the elastic properties (which appear identical). The large elastic improvement on further reaction is shown in later examples.

5g. of polyamide 6 yarn was dissolved in 15g. of 98%
formic acid to form a "dope" which was cast at 22C onto a sheet of high density polyethylene and dried at 60C for 10 hours to give a translucent film which was impermeable to water at 200 kpa at a thickness of 120 microns. The sheet was washed for 48 hours in distilled water and cut to a disc of 45mm diameter. Wedging between metal plates showed a resistance of 200,U00 ohms and only traces of weakly acidic groups, COOH, by methylene blue absorption.
Heating with 25~ weight/volume glutaraldehyde at 100C for 48 hGurs and washing gave a translucent brown disc, with an unchanged resistance of 200,000 ohms but staining an intense purple in Schiff's fuchsin reagent, indicating the presence of many -CHO yroupsO
Heating at 60C in 2% sodium 4 -phenylhydrazinesulphonate and long washing gave a brown disc of lowered electrical resistance, 20,000 ohms showing the presence of conducting ionic groups. Methylene blue then gave an intense blue stain which would not wash out, showing large amounts of SO3 ~ groups. The produc~ was satisfactory for an electrodialysis membrane~ permeable to .

.~ . .
.
., ~ ,'' : , .

~L2~8~i6~

cations. Although ion-exchange properties were shown, the capacity and exchange rates were too low for commercial use. Similar results were obtained using films formed by precipitating a 98% formic acid "dope" by immersion in water. Contrary to the hydrochloric acid "dopes", a porous ultrafilter was not formed, as the 98~ formic acid had dispersed the polyamide 6 molecularly, including the crystallites.

EX~MPL~ ~

To 90g. of the dry polyarnide 6 yarn used in Example 1 were added O.9g of isophthaloy;Lchloride in 180ml of cyclohexane and 3g. of anhydrous potassium carbonate at 22C for 36 hours when 93% of the acid chloride had reacted as determined by the fall in UV absorption at 290 nanometres and the content of chloride reactable with boiling ethanolic silver nitrate in the cyclohexane. The cyclohexane was allowed to evaporate, the fibre washed in water for 1 hour, soaked to pH 3 in dilute HCl, washed overnight and dried at 60C. The isophthaloyl chloride had largely converted some of the amide groups to imide groups with very little -COO~ as determined by comparison of methylene blue absorption with the original yarn.
A "dope" was made up according to Example 1. The dope" was slightly more turbid than that of ExamplQ 1 which indicated some yreater content sf colloidal crystallite or some cross-linking of amorphous polyamide.
The "dope" was cast in parallel with the "dope" of Example 1. ~ comparison of the porous membranes formed showed that the permeabilities to water at 100 kPa for the unmodified polyamide 6 was 117 litre/square metre/hour whereas the imide modified membrane was 97 litre/square metre/hour The polyamide 6/isophthaloylimide-modified membrane above was reacted with glutaraldehyde as in Example 2 with little significant difference from the unmodified polyamide 6 membrane. This similarity extended to the further reaction wit:h xesorcinol according to Example 5. It is ~....... . . :
: : : , . .
.: :

: : . .
~. :

S6~L

clear that the reacting species is the -CO -NH - group and that the imide group -CO - N = is not reactive and merely a diluent whose utility is largely restricted to Eorming a desired physical porous structure. Polyamide 6~6 also reacted as polyamide 6 but was somewhat less resistant to oxidation and to biological attack.

12 gra~s of the glutaraldehyde cross-linked membrane of Example 2 were heated with 12 millilitres of a 1% aqueous resorcinol solution at pH 3.0 at a temperature of 60C for 12 hours and then washed. The resultant membrane had incorporated about 0.2~ of its dry weight of the resorcinol. It was then dyed with p-nitrobenzenediazonium fluoroborate solution. It was not ~oluble in 6N
hydrochloric acid in 1 hour whereas such acid rapidly dissolved precursor membranes. The resultant water permeability of 299 litre/square metre/hour and gelatin rejection of 75% showea that this further modification of the glutaraldehyde membrane occurred without significant change in permeability.
However the ~embrittlem~nt resistance" was raised to a high level - no embrittlement occurred even after 6 m~nths at 60C at pH 7 as against 9 days fcr the embrittlement of the membrane before raaction with resorcinola The ~xtension to break (dry~ was raised from 6~ to 10% and the behaviour to rubbing remaining very rubbery; the extension to break (wet) was raised from 60%
to 70% and the behaviour to rubbing was extremely rubbery whilst the extension to break in ethanol was raised from 20% to 30% with full rubber-like resistance to rubbing.
This membrane made by sequential glutaraldehyde then resorcinol treatment was apparently unaffected by two bio-resistance tests:-a~ Enzymatic. A commercial mixture of papain ~5 and amylase was renewed weekly at 25C to , - : , : .:
: ~ : , . , 6~L

35C for 13 months and a prior art membrane stored therein remained intact but tore easily. In this respect it behaved better than one stored in water since "crystalli~ation embrittlement" was hindered by the contained proteins, presumably because proteins are strongly absorbed and could be expected to hinder crystallization. The treated membrane remained very strong, tcugh and rubbery for the 13 months.
b. Compost Burial. The membranes were soaked in a commercial compost con~aining added commercial "Organic Compo~t Accelerator~ for 13 months at 25C. The untreated polyamide 6 membrane still showed reasonable streny~h but was not comparable to the apparently unchanged tough rubbery nature o~ the Example 5 membrane. A membrane made from polyamide 6,6 rather than polyamide 6 but otherwise treated according to Example~ 2 and 5 showed poor bio-reslstance to compost and easily broke up. Outstandingly good bio-resistance ~to enzymes and compost) was also shown by membranes treated with glutaralaehyde according to ~xample 2 and then:-~a) reacted with hydrazine at pH 5.0 for 10 hours and washed or (b) reacted with excess 2~4-dinitrophenylhydrazine in N/100 HCl overnight, then washed.

The glutaraldehyde/resorcinol treated membranes of Examples 5 are preferred for ultrafiltration purposes and as a stable matrix of controllable porosity from which : , . .
. , , , : .
.:. . , .. :
. . . :
.: ; . : .
: ~, - - ... : . . . .
- . . .. . ..

.- - : . .

' ~ . ~ ' . ': .: ' chemical derivatives for ion-exchange or enzyme immobilization or affinity chromatography can readily be made as described in Example 6.
Repetition of the reactions but substituting hydroquinone, tannic acid or 2 - naphthol - 3,6-disulphonic acid or resorcinol gave analogous products showing expected properties. For example, the tannic acid product formed dark blue-black ferric derivatives; the naphthol-disulphonic acid derivative showed cation-exchange properties. None was physically superior, nor more convenient in ultrafilter manufacture than resorcinol. It is relatively certain that any commercial bisaldehyde/reactive phenol sequence will cross-link and stabilize against "embrittlemant" due to amorphous polyamide recrystallization but glutaraldehyde has overall advantages as a reactant, although further reaction will provide tough, or more rubbery products.

The membrane of Example 5 ~lg) was freed of trace -CHO groups by reaction with dilute hydra~ine at pH 3.5 at 80C for 15 minutes and washed well. The resultant membrane wa~ treated with 0.45g epichlorohydrin in 10ml 95%
ethanol at pH 10 to 12 by adding 0.5ml 2N NaOH at 80C and then washing well. The presence of combined epoxy-groups was demonstrated by slow precipitation of AgI03 on adding AgIo4 in 2N HNO3. The epoxide was reacted with L~ aqueous hexamethylenediamine or 1~ diethylenetriamine by heating to 80C for 30 minutes and then washed well. The presence in both cases of bound -NH2 groups was shown by colorimetric estimation with p~dimethylaminophenylazobenzene-4-isothiocyante. The products were then heated to 80C with excess 1% alcoholic 1,4-phenylenebisisothio-cyanate when the -~H2 groups were converted to the yellow 4-, -, . :
', ' ' ~IL21~

- lg -isothiocyanotophenylthioureas (1). The isothiocyanato- end groups were estimated colorimetrically by reaction with 5-aminofluorescein to give the salmon-coloured derivative.
Throughout the entire sequence the membranes retained their desirable ultrafiltration characteristics. The desirable isothiocyanate intermediates (1) may be regarded as polyamide/(imide)/aldehyde/polyphenol/epoxy/diamine/thio-ureaphenylisothiocyanates. They are dimensionally stable~
controllable-porosity structures with ability to be heat sterilizedO They are especially preferred for reaction with the free -NM2 groups of proteins to give immobilized enzymes or affinity chromatographic column supports.

The membrane of this Example is sterile and ready for use whereas prior art membranes require reaction with a variety of multifunctional reagents before they can be used as is explained in Canadian Patent 1,083,057.

A membrane of this example was treated with an immunoglobulin at 25C with the pH in the range of 7 to 9 to ensure optimal reaction. The binding was unaffected by the addition of sodium chloride of up to 3 Molar. At pH 9 in a 0.01 M phosphate buffered saline solution at 25C, the reaction exhibited diffusion limitations as expected because the membrane pores would not pass the immunoglobulin as its molecular weight of 156~000 is close to the molecular weight cut off of the membrane. The immunoglobulin was applied at 500 microgram/millilitre and the membrane surface took up 512 micrograms/s~uare centimetre. The membrane was then washed to remove non-combined immunoglobulin. Ethanolamine was then added to block off the unreacted isothiocyanate and the membrane was then washed. Incubation with protein and further washing did not lead to takeup of the protein.

, .

-:

..

.. . , . _ The polyamide 6/glutaraldehyde membrane of Example 2 after drying at 60C reacted readily with 0~5% aqueous gelatin, draining, then heating in an oven at 60C for 15 hours. The product was fully "embrittlement resistant~ and had an extension to break of over 50% in absolute ethanol (versus 20~ without gelatin) and was fully rubbery. The membrane showed some utility in rejecting fine oil droplets when used as a cross-flow ultrafilter on oil emulsions in water. Similarly substitution of high molecular weight hydroxyethylcellulose for gelatin gave equivalent membranes which were "embrittl~ment resistant" and rubbery in ethanol with much the same utility in filtering oil emulsions.

The glutaraldehyde in Examples 2,3,5,6 and 7 was replaced with glyoxal, succindialdehyde, phthaldialdehyde and terephthaldehyde. There was little difference in behaviour but the products ~rom terephthaldehyde tended to be too hard for ultrafilters, although the hardness could be turned to useful account when powdered high-pressure liquid affinity chromatographic packings were needed. The aromatic bis-aldehydes tended to be rather slow in reaction but always gave lighter-coloured products. The reactivity Qf all intermediates was in line with the properties of the parent aldehydes eg: polyamide 6/aliphatic bis-aldehydes gave membranes which contained -CHO groups readily reacting with NaHSO3 (stained by Schiff's reagent) whereas the aromatic bis-aldehydes reacted slower. However, all formed 2,4 - dinitro - phenylhydrazones, as expected.
The use of the cheap glyoxal~ glutaraldehyde and terephthaldehyde (if desired by mixing these and, if desired, including a very limited am~unt of formaldehyde) ' :, " ' ' . , S~

can meet all likely needs in serving as a vital intermediate step in the conversion of the desirably structured known polyamide/(imide) membranes into "embrittlement resistant" membranes by further reaction with reactive phenols, proteins or other aldehyde-reactive substances. These can further form desirable derivatives for ultrafiltration~ cross-flow filtration, ion-exchange, protein immobilization or packings for affinity chromatography. The vital point is t~at all of thi~ can be done by immersion in suitable reagents whilst still retaining the carefully controlled initial porous structure.

lOOg. of a 60C dried polyamide 6 based membrane made according to Example 1 and containing 4% of reacted ~lutaraldehyde and 4% of reacted resorcinol based on the dry weight of polyamid membrane was heated 24 hours at 60C with 400ml of a solution of 75ml of 25% weight/volume glutaraldehyde and 40g. of sodium benzoate buffer per litre. The original polyamide/glutaraldehyde reaction product contained only the equivalent of 1% of glutaraldehyde with a reactive single -CHO group as judged by rapid reaction with 2~4-dinitrophenylhydrazine.
Furthermore the product was not stable to dilute acids, gradually releasing more aldehyde. However, by the present ~example it was possible to obtain the equivalent of 2% of single-linked glutaraldehyde which was now linked to a much more stable rubbery matrix. This doubling of the capacity to form derivatives is very important for ion-exchangers and ion-specific ultrafilters, eg: rejecting anionic detergents after treatment with bisulphite.

- ~: . :
- .
'' , ' - '

Claims (31)

1. A polymeric porous membrane comprising a membrane matrix made from an aliphatic thermoplastic polyamide or from an aliphatic thermoplastic polyamide/polyimide copolymer which has both relatively non-crystalline and relatively crystalline portions joined together by relatively non-crystalline portions with pores in the membrane being defined by spaces between the relatively crystalline portions characterised in that at least some of the relatively crystalline portions have been linked together by the reaction of a polyfunctional aldehyde with the membrane matrix.

2. A membrane according to claim 1 wherein the aldehyde is a bis-aldehyde.

3. A membrane according to claim 2 wherein the bis-aldehyde is glutaraldehyde.

4. A membrane according to claim 3 wherein the thermo-plastic material is polyamide 6 and each polyamide chain has a number of amide groups spaced apart along the chain and wherein the glutaraldehyde has displaced the hydrogen atom of the amide group with its end carbon atom becoming bonded to a nitrogen atom in the polyamide chain at least in part as follows:-

5. A membrane according to claim 1 modified in that from 10 to 25% of the aldehyde chains are not linked at each end to a polyamide chain but one end is unattached to leave the CH = O in a more reactive form group.

6. A membrane according to claim 5 wherein the free end of the single-link aldehyde chains have been reacted with a phenol so that the free end of the aldehyde is a bis (phenylol) methane as follows:--CH = O ? -

7. A membrane according to claim 6 wherein the phenol is selected from resorcinol, diphenylolpropane, tannic acid, pyrogallol, hydroquinone, meta-cresol and naphthol as well as derivatives or mixtures thereof.

8. A membrane according to claim 3 modified in that formaldehyde is used as an additional linking reagent.

9. A membrane according to claim a wherein the free ends of the formaldehyde link have been reacted with resorcinol.

10. A membrane according to claim 2, wherein the bis-aldehyde is selected from the group comprising glutaraldehyde, glyoxal, succinic dialdehyde, alpha-hydroxyadipic aldehyde, terephthalic dialdehyde or phthalic dialdehyde or mixtures thereof.

11. A membrane according to any one of claims 1, 4 and 5 wherein the membrane contains reactive aldehyde groups and wherein the membrane has been modified by sodium bisulphite, hydroxylamine - 0 - sulphonic acid or phenylhydrazinesulphonic acid to form a cation exchange membrane.

12. A membrane according to claim 6 wherein the phenol is resorcinol and the membrane was further reacted with sodium monochloroacetate in aqueous solution.

13. A membrane according to claim 6 wherein the membrane wherein free aldehyde groups were reacted with

14. A membrane according to any one of claims 1, 4 and 5 wherein free aldehyde groups are reacted with proteins or polyhydric colloids to form hydrophilic membranes.

15. A membrane according to any one of claims 6, 7 or 9 wherein:-(i) traces of the remaining reactive aldehyde groups have been reacted with hydrazine, (ii) the phenolic hydroxyl groups have been reacted with epichlorohydrin, (iii) the resultant epoxides have been reacted with diamines to fix a pre-determined concentration of amine groups and excess epoxide has been hydrolysed to hydroxyls, and, (iv) the amine groups have been reacted with excess bis(isothiocyanate) to produce a reactive substrate suitable for combining with the free -NH2 groups of biological substances by forming covalent thiourea linkages to provide immobilized proteins or affinity chromatographic media.

16. A membrane according to claim 2 wherein the bis-aldehyde is derived from a bis-aldehyde polymer, an acetal or an acetal ester.

17. A membrane according to claim 5 wherein at least some of the free ends of the single-link aldehyde chain were reacted with gelatin or hydroxyethylcellulose.

18. A method of preparing a porous membrane comprising the steps of:-(i) dissolving an aliphatic thermoplastic polyamide or an aliphatic thermoplastic polyamide/polyimide copolymer which has both relatively non-crystalline and relatively crystalline portions into a solvent under conditions of temperature and time which cause the relatively non-crystalline portions of the polyamide or copolymer to dissolve while at least a part of the relatively crystalline portions of the polyamide or copolymer do not dissolve, but form a colloidal dispersion in said solvent, (ii) forming said colloidal dispersion into a film and thereafter causing precipitation of at least part of the dissolved non-crystalline portions in the film to form a porous membrane matrix in which the pores are defined by spaces between the relatively crystalline portions, and, (iii) reacting the membrane matrix with a polyfunctional aldehyde to link at least some of the relatively crystalline portions with the aldehyde.

19. A method according to claim 18 wherein the aldehyde is a bis-aldehyde.

20. A method according to claim 19, wherein the aldehyde is glutaraldehyde, glyoxal, succinic dialdehyde, alpha-hydroxyadipic aldehyde, terephthalic dialdehyde, phthalic dialdehyde and mixtures thereof.

21. A method according to claim 18 wherein the aldehyde is derived from a bis-aldehyde polymer, an acetal or an acetal ester.

22. A method according to claim 19 wherein aldehyde reaction step is so controlled that from 10% to 25%
of the aldehyde chains are not linked at one end.

23. A method according to claim 22 including the step of reacting at least some of the free ends of the single-link aldehyde chains with a phenol.

24. A method according to claim 23 wherein the phenol is selected from the group comprising resorcinol, diphenylol propane, tannic acid, pyrogallol, hydroquinone, meta-cresol and naphthol as well as derivatives or mixtures thereof.

25. A method according to claim 22 including the step of reacting at least some of the free ends of the single-link chains with a protein or polyhydric colloid.

26. A method according to claim 22 including the steps of reacting at least some of the free ends of the single-link chains with gelatin or hydroxyethyl cellulose.

27. A method according to claim 23 including the step of reacting the phenol modified chain with sodium mono-chloroacetate in aqueous solution.

28. A method according to claim 23 including the step of reacting the phenol modified chain with aqueous diazonium salts.

29. A method according to claim 23 including the steps of:-(a) reacting at least some of the remaining reactive single-link aldehyde chains with hydrazine, (b) reacting the phenolic hydroxyl groups with epichlorohydrin, (c) reacting the resultant epoxides with diamine to fix a pre-determined concentration of amine groups and hydrolysing excess epoxide to hydroxyls, and, (d) reacting the amine groups with excess bis (isothiocyanate).

30. A method according to claim 23 and including the step of reacting the phenol modified membrane with a bis-aldehyde.

31. A method according to claim 18 wherein the membrane is reacted with sodium bisulphite, hydroxylamine-o-sulphonic acid or phenylhydrazinesulphonic acid.