CA1101827A - Bimodal chromatographic resolving zone - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE

Disclosed herein is a resolving zone for a chromatograph comprising a plurality of porous, silica macroparticles chosen to provide the resolving zone with a bimodal pore distribution, the average pore size for each mode being such that the linear portions of the molecular weight calibration curve for each pore size in the bimodal distribution are nonoverlapping and the pore volume of each mode being such that the aforesaid linear portions are substantially parallel. The macro-particles can be either totally or superficially porous.
The resolving zone can be composed of either a plurality of macroparticles each having a bimodal pore distribution or a plurality of macroparticles having one pore size distribution and a plurality of macroparticles having another pore distribution.

Description

BACKGROU~D OF THE INVENTION
1. F:ield of the Invention: This invention relates to chromatography, particularly size exclusion chromato-graphy, and to the composition of the resolving zones used in such chromatography. It also relates to a process for performing chromatographic separation.

2. Discussion of the Prior Ar_: U S. Patent

3,505,785 discloses superficially porous microspheroids, having an average diameter in the range of 5 to 500 microns, ].0 which are composed of an impervious core coated wlth a multiplicity of monolayers of colloidal inorganic particles having an average size in the range of 0.005 ~o 1.0 microns.
U.S. Patent 3,855,172 discloses microspheroids which are ~ s~
porous throughout and have an average diameter in the range of 0.5 to 20 microns. They are composed of colloidal in-organic refractory particles having an average diameter in the range of 0.005 to 1.0 microns.
The use of such microspheroids in chromatography, particularly size-exclusion chromatogxaphy is well-known.
In such application, porous microspheroids are used as `
packings for the chromatograph resolving zone and functions to separate the components of a sample based on differences in hydrodynamic size of the components~ The molecular weight (MW) of the components can be calculated as a function of the hydrodynamic size. A plot of the molecular weight ;
fraction eluted at a given retention volume (VR) for a 1~7 particular packing material reveals that for a given pore volume in the packing material, certain molecular weight fractions are totally excluded because of theix large size and certain mol~cular weight fractions are to~ally permeating because of their small size. Between these two extr~mes i~
a range of molecular weight fractions that w;ill be prefer-entially retarded by contact with the porous particles, and matexials containing these molecular weight fractions can be fractionated by that particular packing material.
The actual working relationship in size exclusion chromatography is the molecular weight calibration curve which is normally a logarithmic plot of the molecular weight versus the retention volume. Molecular weight calibration curves characteristically have a substantially linear portion so that the molecular weight of a retained fraction of a sample can be determined accurately if the retention volume for that particular molecular weight fraction occurs in the linear portion of the molecular weight calibration curve and less accurately if it occurs outside that linear range.
Molecular weight calibration curves characteristically have linear portions that span approximately two decades in the log molecular weight scale for a single pore size. To obtain a calibration curve ~ith a linear portion spanning more than two decades of the molecular weight scale, the tendency is to use a chromatographic resolving æone composed of many cOlUmn5, each having different pore sizes. Specifically, as taught by "Know More About Your Polymeri', a 1976 pu~blication of ~aters Associates, Milford,Mass., to expand the linear rang~ of the molecular weight calibration cuxve so that materials containing a wide range of molecular weight xactions can be separated and detected, four or five columns are combined, each with molecular weight cali-bration curves whose linear portions o~erlap one another.
Unfortunately, the range of expected linear molecular weight calibration does not occur~ The linear portion of the molecular weight calibration curve ~or the combined particles can be increased in this way but the maximum appears to be only about three decades, which often is less than the range of molecular weight found in normal sample ~ ~`
compositions.

SUMMARY OF THE INVENTION
According to this invenkion, there is provided a resolving zone for a chromatograph which provides a wider range of linearity in its molecular weight calibration curve and avoids the inconvenience of using multiple columns with overlapping linear calibration. This resolving zone comprises a pLurality of porous macroparticles chosen to provide the resolving zone with a bimodal pore distribution, the average pore size ~or each mode being such that the linear portion of the molecular weight calibration curves for each pore size in the bimodal distribution are substantially nonover-lapping, the pore volume of each mode being such that the aforesaid linear portions are substantially parallel. As used herein, the term "average pore size" means volume average pore size. To achieve the maximum range of the linear por-tion of the molecular weight calibration curve, the pore sizes of the modes of the bimodal distribution shoul~ be about one order of magnitude apart. The term "bimodal pore distri-bution" as used herein is not meant to exclude the use of more 30than two modes o~ particle size distribution so long as two adjacent modes have substantially parallel linear non-overlapping portions of their cal.ibration curves, the average pore volume of each of the two adjacent modes being separated about one order of magnitude.
The macroparticles useful in this :invention can be refractory particles such as silica o:r alumina or they can be non-refractory such as crossl.inked polymer gels. In the preferred embodiment the macroparticles are refractory macroparticles composed primarily of silica and the component of the bimodal pore distribution ;
having the smaller average pore size provides about 30 to about 60%, preferably about 40 to about 60%, more pre~erably about 40 to 55% and still more preferably, about 45 to 55%, of the total pore volume of the macro-particles in the resolving zone with the balance of the total pore volume being provided by the component . :
of the bimodal distribution having the larger average pore size. The resolving zone can be composed of a plurality of macroparticles, each having a bimodal pore distribution, or it can be composed of a plurality of ~ .
macroparticles each having one pore distribution com- :
bined with a plurality of macroparticles having another pore distribution.
This invention also provides an improved process ~or performing chromatographic separation comprising the steps of:
(a) placing the material to be separated in a carrier fluid;
(b) contacting the carrier fluid with a resolving zone comprising a plurality of porous, refractory, macroparticles chosen to provide the resolving zone with a bimodal pore distribution the average pore si.ze for each mode being such that the linear portions of the molecular weight calibration curve for ~ach pore size i.n the bimodal distribution are nonoverlapping, and -the pore volume of each mode being such that the aoresaid linear portions are substantially parallel, and ~ c) determining the extent of retention o~ the materials by the resolving ~one.
BRIEF DESCRIPTION OF TH~ ~R~WINGS
The present invention can best be described by reference to the following figures in which:
Figure 1 is a representatiVe calibration cuxve of molecular weight plotted on a log scale ag~inst retention volume;
Figure 2 is a calibration curve for resolving zone composed of six different particles having differing pore sizes; ;
Figure 3 is a schematic representation of a liquid chromatograph showing the carrier fluid injection point 12, resolving zone ~3 ~nd detector 14; showing particularly resolving zone~ composed o~ 1 and 2 separate columns;
Figure 4 is a represent~tive calibration cuxve for a bimodal and a polymodal resolving zone;
Fi~ure 5 is a partially cut-away schematic ;
representation of one embodiment of a totally porous ;~
macroparticle having a bimodal pore distribution;
Figure 6 is a partially cut-away schematic representation of a second embodiment of a totally porous 3~ macroparticle w~th a bimodal pore distribution;

Figure 7 is a partially cut-away schematic representation of an embodiment of a superficially porous macroparticle having a bimodal pore distribution;
Figure 8 is a mercury intrusion plot for particles such as that shown in Figure 6;
Figure 9 is a comparison o a calibration curve obtained using two of the particles, which had nonoverlapping particle size distributions, shown in Figure 2 compared with a calibration curve obtained using five of the particles all ~;

having overlapping particl* size distributions shown in Figure 2; and Figure 10 is a calibration curve obtained using the bimodal particles such as that shown in Figure 6.
DETAILED DESCRIPTION OF T~IE INVENTION
In size exclusion chromatography, a chromatograph such as that shown schematically in Figure 3 is used. A
material to be separated is injected into a carrier fluid stream at some injection point 12 and forced under pressure through a chromatographic resolving zone 13 to a detector 14. In passing through the resolving zone, the materials in the carrier fluid contact the packing material in the resolv-ing zone and are retained for a time characteristic of their molecular weight ~MW). In time, as more volume of carrier ~luid passes through the chromatographic column, the material temporarily retained by th~ packing material is eluted from the column. The detector determines when each component of the material leaves the resolving zone.l The output of the detector is characteristically a peak such as that shown ~chematically in the bottom of Figure 4.

Th~ resolving zones used in size exclusion ~'.

-7~

chromRtography are gen~rally column~ paeked with porou~
partlcle~ ~uch a~ tho~e de~cribed ln U~ S~ :Pat~nt 39 505,785 and U. S. Pat~nt 3,,855,1729 or more recently the macroporou~
micro~pheroid~ di~clo~e~ ~n U.S~ en~ 4,o7~o~,~86 of }~. K.
Iler and J. J,, ~irkland"

U~ing æuch packing mat~rial3, a typlcal relatlc~n~
~hlp o~ th~ lo~ a funcltlon of ~olute hyclrodynamlc rad~us versus retentiDn vo~Lume (VR) is ffhow~ by th~ single ~olld line on the left-hand a lde of Figure 1, The liLmi ting reten-10 tion volume at a p~31nt A 1~ known a~ the total ex~luslon volum~, which i~ determined by the maximum pore ~ize avail ~.
able for permeation by the ~olute material~ that ar~ totally re~ect~d ~rom the ln~ernal poro~ity ~luts at thls retentlon volume, an~ 801ute~ corr~spon~i~g to this molecular w~ght and l~rger are not fraction~ted by the ~y~tem. Point B ~ ;
repre~e~ts th~ ~rolume a~oc1ated wlth ~pec~es which totally ::
permeate the lnternal porea o~ the packlng materlal3 and i3 known ~ the to~al perD~eat~on volume. Thus, matel~ial~
corresponding t~ this MW and 3maller can~c~t be ~ub~tankillly 20 ~raction~ted by lihi~ ~eparatlng 8g8t~m~ The dl~erencebet~een retention volume~ A and B repres~nt~ partial permea-tiion o~ ~olute~, and it i~ wlthin t~ volume rallg~ that ~epar~tion ~3ccur~. The difrerezlc2 be~ween the retention volume at B ~nd re~entir:,n vol~me at A 18 a f~nc~ion o~ the total intern~l pore volume o~ the packiIIg. B~tw~en ret~ntion volume~ ~t A and B th~rs i~ an approxlmately lin~ar region of th@ log MW ~er~u~ retent on Yolum~ curve (point~ C to D) whlch 1~ described by th~ rollowing ~quat;ion~: :

.~

v~ a cl -- c2 lQgMW (l ) and MW - Dle~D2VR (2) C2 is ~he slope of the linear portion of the calibratior curve (in ml/decade-MW~ and Cl is the intercept of thls `
linear portion. To extract molecular weight information from this calibration plot, experimental chromatograms and equation 2 above are utilized. Dl relates to the intercept of this linear portion of the calibration curve and ~ relates `~
to its slope. These equations are well-known to those skilled in the art and are widely used by those characterizing ~;
macromolecules~
The additi~e characteristics o~ two identical columns used in size-exclusion chromatography is well known.
As indicated in Figure 1, connecting two identical columns ~same particles, same length) with identical cali.bration curves indicated by the two solid lines is equivalent to doubling the length of a single column. As indicated, connecting these two columns increases the total available pore volume, thus increasing the retention volume range ' between total permeation and total exclusion, but maintaining the same molecular weight fraction range. As shown in Figure 1 when two columns are connected, the molecular weight fraction range remains about 3,000 to 80,000, even though the retention volume is doubled. The calibration curve for the combination of the two columns is shown by the dashed line. The additive function describing this relationship is:

C2 = i (C2)i~ or (3) D2 = 1/~ (1/D2)i (4) -Traditionally, polymer fractionation has been _g_ :

.:; ,: . .

2~ `

accomplished with packings having the broadest possible pore-size distribution. This is normally obtained by connecting several columns o different pore size to produce a separating system covering the molecular weight range of interest. Figure 2 shows a series of molecular weight cali-bration curves for six different chromatographic resolving zone, each filled with porous silica particles having differ-ent pore sizes. ~he designation and average pore volume f~r these six particles is given in Table I below:

TABLE I
Designa-tion Pore Size (A)

4 PMS-800 300 These particles were made as described in U.S. Patents 3,782,075 and 4,070,286.
The bar graphs to the right of Figure 2 indicate the linear range of each calibration plot. To achieve a linear combined calibration curve spanning a molecular weight range from 103 to 106, a combination of six columns traditionally would be used r each composed of the individual particles corresponding to the six graphs.
In this invention the rela-tionship given in equation 3 has been exploited to improve the accuracy, ver~atility, and convenience of the size-exclusion process. This rela-tionship predicts a previously unrecognized phenomena, namely, that to obtain a wide linear log MW-retention volume rela~
tionship, a series of columns having substantially overlapping linear molecular weight fractionation ranges (i.e., llnear portions) should not be used. Rather, columns having only two pore sizes, chosen so that the linear portions of the -molecular weight versus retention volume curves do not overlap, should be used. This produces a far wider linear range in the calibration curve. As shown representatively in Figure 4, a polymodal pore distribution in the resolving zone produces a narrow linear portion on the molecular weight calibration curve, and a bimodal distribution pro-duces a much wider linear portion. The calibration curve for the polymodal distribution does not encompass the entire molecular weight distribution of the sample within its linear ranger whereas the calibration curve for the !, bimodal distribution does.
The advantage of using chromatographic columns having a bimodal pore distribution, whether connecting columns of individual pore size or using columns containing a physical mixture of particles that are two pore sizes can, therefore, be seen from Figure 4. Molecular weight calibration curves of the type shown by the bimodal pore size distribution is greatly preferred when attempting to characterize a polymer with the type of molecular weight distribution illustrated at the bottom of the plot.
A quantitative comparison of a polymodal pore-size distribution versus bimodal pore-distribution systemlis given in Figure 9. Figure 9 shows a polystyrene calibration curve for a polymodal and for a bimodal resolving zone.
The set of columns used to produce the polymodal distrihution -11 ;

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

are filled with a packing material labeled l, 2, 4, 5 and 6 in Figure 2. The individual columns have substantially overlapping calibration plots as in the traditional mode.
The appro~imate linear calibration ranqe tdashed line is the linear fit) of this combined broad pore-size distribution set is only about two and one-half decades of molecular weight. On the other hand, the bimodal distribution shown in Figure 9, obtained by connecting columns of only two pore sizes (that of particles l and 5 in Figure 2) results in a linear molecular weight calibration curve spanning more than four decades of molecular weight.
To obtain these unexpected and improved results, the individual calibration curves for the two pore sizes used in the bimodal distribution must not overlap. This is achieved by choosing particles with the appropriate pore size. The average pore sizes of the bimodal distribution should be about one order of magnitude apart. With this bimodal approach, linear calibration curves having up to five decades of molecular weight range ar~ obtained. A
trimodal arrangement of similar type could result in up to seven decades molecular weight range linearity. In addition to having molecular weight calibration curves which are non-overlapping, the internal pore volume of the two modes should be such that the linear portions of the calibration curves are substantially parallel. The term substantially parallel means that the shapes of the linear portions of the calibra-tion curves need not be exactly paraLlel provided sbme devi-ation from linearity can be accepted. For example, reference to Figure 9 indicates that the overlapping polymodal calibra-tion curve which represents the prior art is far from :Linear over the range predicted. When about 30 to 60~, pre-ferably 40 to about 60%, more preferably about 40 to 55%
and still more preferably about 45 to 55~, of the total pore volume of the macroparticles in each resolving zone is provided by the component of the bimodal pore distri-bution having the lower average pore size with the balance provided by the component having the larger average pore size, the linear portions of the individual calibration curves are substantially parallel. Reference to Figure 10 shows the deviation from linearity in the calibration curve when the pore volume ratio is 40:60.
In the most preferred embodiment, however, each component -of the bimodal pore distribution should provide about 50% of the total pore volume of the macroparticle in the resolving zone to reduce deviation from linearity in the calibration curve.
Best results are obtained using packing materials with a very narrow pore size distribution.
Pore size distributions of each of the bimodal systems should be 1.0 or ~ess (2a) as shown in conventional log-normal plots of mercury porosimetry measurements.
Ranges of 0.5 (2a) are preferred. Withln these ranges, pore size distribution is an insignificant factor in determining the D2 of the calibration plot, and the internal volume of the particles is dominant in deter~
~ining the D2 of the calibration plot. If the pore size distribution is larger than the values given a~ove, then the interrelationship of both pore size distribu-tion plus internal volume determines the slope of the calibration curvP.

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

The bimodal pore distribution used in the present invention can be achieved in one of two ways. The bimodal pore distribution can be provided by a plurality of micro~
particles each having a bimodal pore distribution. In this instance, a single column such as that sho~l in the upper portion of Figure 3 can be used. Alternately the bimodal pore distribution can be provided by using a plurality of macroparticles having one pore distribution and a plurality of macroparticles having another pore distribution. While particles with different pore size distributions can be mixed into one column, the packing of such columns is less convenient and it is best -to use two or more columns, each packed with a single type particle.
Individual particles of the desired pore size to produce the bimodal pore distribution can be produced by the -techniques described in the patents and the patent applications mentioned above. Polymeric gels, alumina and the wide range of refractory particles mentioned in these documents can be used, but silica is the preferred material, particularly for chromatographic separations. Paxticles having a bimodal pore distribution can either be totally porous or superficially porous macroparticles. The term macroparticle, as used hereîn, means the composite macro-particle (either totally or superficially porous) having an average diameter in the rangé of about 0.5 to about 500 microns. The totally porous em~odiment of this par~icle is shown in Figure 5. ~ere the macroparticle 15 has an average diameter of about 0~5 to S00 microns. Preferred macroparticles ,, .

. . . ~, ' havinq averac~e diameters of about 5 to 50 microns. The macroparticle is composed of a plurality of microparticles 16, each having an average diameter in the range of about 0.005 to ~bout 1.0, preferably abou~ 0.005 to 0.5/ microns.
The individual microparticles are in turn composed of a plurality of ultramicroparticles 17 having an average dia-meter in the range of about 1.0 to about 30.0 nanometers with 2-20 being preferred. Between each microparticIe is a macro-pore 18, and between each ultramicroparticle is a micropore 19. While these particles can in qeneral have any shape, it is preferred that they have a spherical shape so that the macroparticles are actually macrospheres, the microparticles are microspheres and the ultramicroparticles are ultramicro-spheres. The spherical nature of these materials improves their performance in chromatographic columns.
Alternatively, as shown in Figure 6, the totally porous macroparticle 15 can be composed of a core 20 compris-ing a plurality o~ ultramicroparticle 21 having an average diameter in the range of about 1 to about 30 nanometers, and ~ `
a skin composed of a multiplicity of microparticles 22, each having a diameter in the range of about 0.1 to about 1.0 microns, or moxe commonly, 0.1 to 0~5 microns. The totally porous macroparticle produced by currently known techniques ;~
preerably have a diameter in the range of about 0.5 to about 50 microns.
One embodiment of a superficially porous macropar-ticle is shown in Figure 7. Such macroparticle 22 hlas a diameter in the range of about 0.5 to about 500 or preferably,

5 to 50, microns and comprise an imp~rvious macrocore~24 and a coating of a multiplicity of like monolayers of like colloidal -15~

- ~- , , , ~

inorganlc microparticles 25 joined to and surrounding the core. Each microparticle has an average diameter in the range of about 0.005 to aoout l.0 microns or preferably, 0.1 to 0.5 microns and comprises from about 0.2 to about 25% of the total volume of the macroparticle. The microparticle can be similar to that shown in Figure 5, composed totally of ultramicroparticle, or it can be similar to the microparticle shown in Figure 7, composed of an impervious microcore 27 and a coating of a multiplicity of like monolayers of like colloi-dal inorganic ultramicroparticles 28 joined to and surrounding the core.
In either case, for the totally porous or the super-ficially porous particles, the pores between the individual microparticles in the macroparticle shall be referred to as th~ macropore.30 and provides one mode of the bimodal pore distribution, and the pores between the individual ultramicroparticle shall be referred to as the micropore 31 and provides the other mode of the bimodal pore distribution.
Recent terminology sometimes defines pores of the size designated as "micropores" herein as ~Imesopores~.

E~AMPLE I
The following describes the preparation of pelli-cular particles with a bimodal pore-size distribution. Such a structure is shown in Figure 7~
75 g of Zipax~ controlled porosity support (E. I.
du Pont de Nemours and CoO) (<37~m~ was stirred gently with 800 ml of 0.5% Lakeseal laboratory cleaner solution~for 30 Zipax~ - Du Pont trademark for chromatographic support Z7 ~

minutes. The excess solution was removed by decantation and washed with distilled wa~er. This operation was repeated seven times, and the resulting powder filtexed on a coarse sintered-glass filter and dried in air. The dry powder was then placed in a three-inch diameter coarse sintered-glass funnel and treated with lO0 ml of 0.5% Zelec~ DX (E. I. du Pont de Nemours and Co.) solution for five minutes with stirring. The treated beads were filtered, then washed twice with 200 ml o distilled water and dried in the funnel with vacuum-The beads then were treated with lO0 ml o lO~
silica sol made from Ludox~ AS (~140A silica particles supplied by E. I. du Pont de Nemours and Co.) (125 g of 30% by weight silica in Ludox~ AS diluted to 400 g with distilled water).
The mixture of beads and silica sol was allowed to stand for l5 minutes in the funnel with frequent gentle stirring.
Excess Ludox~ was then filtered off and the resulting wet ;
cake washed four times by gently slurrying with about 400 ml of tap water and filtering. The cake was then allowed to ;
air-dry in the filter under vacuum. This material was then dried at 150C for one hour in a circulating air oven and a small sample removed for surface area measurement.
The Zelec~ DX silica sol treatment described above was repeated three more times to build up a crust of the 140A silica sol ultramicroparticles on the surface of the Zeleo~ - Du Pont trademark for antistatic agents and mold release agents Ludox~ - Du Pont trademark for colloidal silica , . . ' . : - , ~ :. ~ .

2000A silica mi~roparticles which originally made up the crust of the Zipax~ particles. The final particles were dried, and haated at 650C for two hours to burn out the organic interlayer and sinter the particles into a mechAni-cally stable condition. This sintered sample was then allowed to stand for two hours in a large excess of 0.001 M
ammonium hydroxide with frequent stirring. The particles were then washed twice with a large excsss of distilled water by decantation, filtered on coarse sintered-glass funnel, air dried, and heated at 150C for two hours in a circulaking air bath. The final material was dry-sieved with stainless screens to obtain a <38~1m fraction of 45 g.
Surface areas on the products obtained during the synthetic steps were obtained by the nitrogen flow method with the following results:
Sample Surface Area, m2/g _ _ _ _ Starting Zipax~ 0.89, 0.99 First treatment with Ludox~ AS 2.03, 2.03 Second treatment with Ludox~ AS 2.35, 2.46 Third treatment with Ludox~ AS 3.07, 3.01 Fourth treatment with Ludox~ AS 3.38, 3.50 Sintered at 650C for two hours 2.67, 2.67 Final rehydrated material 2.85, 2.86 A mercury porosimetry measurement of this sample showed three breaks in the mercury 1ntrusion plot, one at about 10 microns, representing the intrusion of mercury between the individuai particles, a break at about 0.07 (700A3 representing the macropores between the sol micro-particles in the crust of the initial Zipax~ structura, and a break at about 0.006~ (60A) representing the pores between the 140A sol ultramiGroparticles which are mul~ilayered onto the original Zipax~ structure by the procedure herein described. The volumes associated with the bimodal pore-size distribution were:
Macropores - (700A pores) 0.011 cc/g Micropores - t60~ pores) - 0.014 cc/g.
These data show that the final particles contained the dasired bimodal pore configuration, with pores approximately one decade in size difference, and approximately equal pore volumes for each pore size.
EXAMPLE II
Particles of the type illustrated in Figure 6 can be prepared as follows: 15 g of porous silica microspheres (PSM-40; 4~ angstrom pores) made according to U.S.
Patent 3,78Z,075, January 1, 1974, Joseph J. Kirkland, assigned to Du Pont; U.S. Patent 3,855,172, December 17, 1974, Ralph K. Iler and Herbert ~. McQueston, assigned to E. I. du Pont de Nemours and Co., was treated with 200 ml of O.001 M ammonium hydroY~ide, allowed to stand for 10 minutes with occasional stirring and centrifuged in a 250 ml poly- -ethylene bottle for two minutes (from the start) at approxi-mately 2,000 xevolutions/min. The clear supernatant was ~ -decanted, and to the wet cake was added 100 ml of 0.5% Zelec~
DX (E. I. du Pont de Nemours and Co.) solution which~had been adjusted to pH 7 with ammonium hydroxide. The sytem was ~-carefully slurried, then left to stand or 10 minutes with occasional gentle stirring. The resulting mixture was , centrifuged for one minute using the approach described above, and the excess Zelec~ DX solution decanted.
The wet cake was washed twice with 200 ml of distilled water (adjusted to pH 7 with ammonium hydroxide) by carefully slurrying, centrifuging for one minute, and decanting.
To these treated particles was added 50 ml of 5%
(by weight) 2000A silica sol mixture adiusted to pH 8 ~sol can be prepared by procedures in: W. Stober, A. Fink and E. Bohn, J. Colloid, Inter. Sci., 26, 62 ~1968)~, and the mixture was thoroughly slurried and occasionally stirred for lO minutes. This mixture was centrifuged as above, decanted and excess sol retained. The coated beads were then washed twice with 200 ml of distilled water (pH 7) by slurrying, centrifuging and decanting. The second wash decantant from this process was clear. The wet cake was then filtered on a 3~m "Nuclepore" filter and dried in a circulating air oven at 150C for two hours. The sample was then fired at 700C
for one hour in a muffle furnace. A small portion o this material was subjected to scanning electron micron analysis which showed an excellent coverage of the surface of the original beads with the 2000A silica sol. ~o bare spots were seen on ~he particles.
A second layer of 2000~ silica sol was placed on the PSM particles using the technique described abo~e, after first hydrolyzing the fired silica particles in 0.01 M hydro-chloric acid overnight and eliminating the acid by washing with distilled water. The material was again treated with Zelec~ DX, followed hy ~OOOA sol ~25 ml of "virgin" plus the recovered sol excess from the first trea~ment)~ in the manner described above. Inspection of this material by scanning electron microscopy showed the second coating was layered as desired~ Very few spot5 were seen on the beads, and only a very small amount of particle bridging was noted.
A third, fourth, fifth, and sixth treatment o~
the beads were carried out in essentially the same ~anner as described above to build up the desired crust of 20QOA
silica sol particles on the particles. These treated beads were fired at 750C for one hour and rehydrolyzed by dilute acid treatment as above. SEM inspec~ion of the final beads showed good coverage, but it was not possible to observe the exact thickness of the desired superficially porous crust.
Mercury intrusion measurements plotted in E~igure 8 show that the pore volume of the larger pores of these particles is about 40~ of the total pore volume and the pore volume of the smaller pore volume is about 60% of the total.
The log molecular weight versus retention volume calibration plot for a 25 x 0.62 cm i.d. column of these particles i~
shown in Figure 10. Because of the difference in internal volumes associated with the two modes; there is some de~iation ~ -~
from linearity.

I
:.

Claims (41)

1. A resolving zone for a chromatograph comprising a plurality of macroparticles, said macroparticles being chosen to provide said resolving zone with a bimodal pore size distribution, the average pore size for each mode being such that the linear portions of the molecular weight calibration curve for each pore size in the bimodal distribution are substantially nonoverlapping and the pore volume of each mode being such that said linear portions are substantially parallel.

2. The resolving zone of Claim 1 wherein the component of the bimodal pore size distribution having a smaller average pore size provides from about 30 to 60%
of the total pore volume and the component of the bimodal pore size distribution having a larger average pore size provides from about 70 to 40% of the total pore volume of the macroparticles in the resolving zone.

3. The resolving zone of Claim 2 wherein each component of the bimodal pore size distribution provides about 40 to 60% of the total pore volume of the macro-particles in the resolving zone and wherein the average pore sizes of the components of the bimodal pore size distribution are about one order of magnitude apart.

4. The resolving zone of Claim 3 wherein the component of the bimodal pore size distribution having the smaller average pore size provides about 40 to 55% of the total pore volume and the component of the bimodal pore size distribution having the larger average pore size provides about 45 to 60% of the total pore volume of the macroparticles in the resolving zone.

5. The resolving zone of Claim 4 wherein each component of the bimodal pore size distribution provides about 45-55% of the total pore volume of the macroparticles in the resolving zone.

6. The resolving zone of Claim 4 wherein the macroparticles have an average diameter of about 0.5 to 500 microns and are composed of a plurality of microparticles having a diameter of about 0.005 to about 1.0 micron and wherein each macroparticle has a bimodal pore size distribution.

7. The resolving zone of Claim 4 wherein the macroparticles have an average diameter of about 0.5 to 500 microns and are composed of a plurality of micro-particles having a diameter of about 0.005 to about 1.0 micron, the bimodal pore size distribution in the zone being provided by a plurality of macroparticles having an average pore size within one mode of the bimodal dis-tribution and a plurality of macroparticles having an average pore size within the other mode of the bimodal distribution.

8. The resolving zone of Claim 6 wherein said macroparticles are totally porous macroparticles having an average diameter of about 0.5 to 50 microns and are composed of a plurality of microparticles having an average diameter of 0.005 to 0.5 micron.

9. The resolving zone of Claim 8 wherein the macroparticles are silica.

10. The resolving zone of Claim 6 wherein the macroparticles are superficially porous having an average diameter of about 5 to 50 microns and are composed of a macrocore surrounded by microparticles having a diameter of 0.1 to 0.5 micron.

11. The resolving zone of Claim 10 wherein the macroparticles are composed primarily of silica.

12. The resolving zone of Claim 7 wherein the macroparticles are totally porous macroparticles having an average diameter of about 0.5 to 50 microns and are composed of a plurality of microparticles having an average diameter of about 0.005 to 0.5 micron.

13. The resolving zone of Claim 12 wherein the macroparticles are silica.

14. The resolving zone of Claim 7 wherein the macroparticles are superficially porous having an average diameter of about 5 to 50 microns and are composed of a macrocore surrounded by microparticles having a diameter of about 0.1 to 0.5 micron.

15. The resolving zone of Claim 14 wherein the macroparticles are composed primarily of silica.

16. In a process for chromatographic separation comprising the steps (a) placing the material to be separated in a carrier fluid;
(b) contacting the carrier fluid with a resolving zone; and (c) determining the extent of retention of said material in the zone, the improvement comprising using a resolving zone com-prising a plurality of macroparticles, said macroparticles being chosen to provide said resolving zone with a bimodal pore size distribution, the average pore size for each mode being such that the linear portions of the molecular weight calibration curve for each pore size in the bimodal distribution are substantially nonoverlapping and the pore volume of each mode being such that said linear portions are substantially parallel.

17. The process of Claim 16 wherein the component of the bimodal pore size distribution having a smaller average pore size provides from about 30 to 60% of the total pore volume and the component of the bimodal pore size distribution having a larger average pore size provides from about 70 to 40% of the total pore volume of the macroparticles in the resolving zone.

18. The process of Claim 17 wherein each component of the bimodal pore size distribution provides about 40 to 60% of the total pore volume of the macro-particles in the resolving zone and wherein the average pore sizes of the components of the bimodal pore distribution are about one order of magnitude apart.

19. The process of Claim 18 wherein the component of the bimodal pore size distribution having the smaller average pore size provides about 40 to 55%
of the total pore volume and the component of the bimodal pore size distribution having the larger average pore size provides about 45 to 60% of the total pore volume of the macroparticles in the resolving zone.

20. The process of Claim 19 wherein each component of the bimodal pore size distribution provides about 45-55% of the total pore volume of the macro-particles in the resolving zone.

21. The process of Claim 19 wherein the.
macroparticles have an average diameter of about 0.5 to 500 microns and are composed of a plurality of micro-particles having a diameter of about 0.005 to about 1.0 micron and wherein each macroparticle has a bimodal pore size distribution.

22. The process of Claim 19 wherein the macroparticles have an average diameter of about 0.5 to 500 microns and are composed of a plurality of micro-particles having a diameter of about 0.005 to about 1.0 micron, the bimodal pore size distribution in the zone being provided by a plurality of macroparticles having an average pore size within one mode of the bimodal distribution and a plurality of macroparticles having an average pore size within the other mode of the bimodal distribution.

23. The process of Claim 21 wherein said macroparticles are totally porous macroparticles having an average diameter of about 0.5 to 50 microns and are composed of a plurality of macroparticles having an average diameter of 0.005 to 0.5 micron.

24. The process of Claim 21 wherein the macroparticles are superficially porous having an average diameter of about 5 to 50 microns and are com-posed of a macrocore surrounded by microparticles having a diameter of 0.1 to 0.5 micron.

25. The process of Claim 22 wherein the macroparticles are totally porous macroparticles having an average diameter of about 0.5 to 50 microns and are composed of a plurality of microparticles having an average diameter of about 0.005 to 0.5 micron.

26. The process of Claim 22 wherein the macroparticles are superficially porous having an average diameter of about 5 to 50 microns and are composed of a macrocore surrounded by microparticles having a diameter of about 0.1 to 0.5 micron.

27. A powder for chromatographic separa-tions consisting essentially of a plurality of discrete porous macroparticles, each macroparticle having an average diameter of about 0.5 to about 500 microns and a bimodal pore size distribution, the average pore sizes of the components of the bimodal distribution being about one order of magnitude apart and the com-ponent of the bimodal distribution having the smaller average pore size provides from about 30 to 60% of the total pore volume and the component of the bimodal distribution having the larger average pore size pro-vides from about 70 to 40% of the total pore volume.

28. The powder of Claim 27 wherein the component of the bimodal pore size distribution having the smaller average pore size provides about 40 to 55%
of the total pore volume and the component of the bimodal pore size distribution having the larger average pore size provides from about 60 to 45% of the total pore volume.

29. The powder of Claim 28 wherein each component of the bimodal pore size distribution provides from about 45 to about 55% of the total pore volume.

30. The powder of Claim 27 wherein said macroparticles are superficially porous macroparticles having an impervious core and a coating of a multi-plicity of like monolayers of colloidal microparticles joined to and surrounding the core, the microparticles having a diameter of about 0.005 to about 1.0 micron and comprising from about 0.2 to about 25% of the total volume of the macroparticle.

31. The powder of Claim 30 wherein the microparticles have a diameter of about 0.1 to 0.5 micron and the macroparticles have a diameter of 5 to 50 microns.

32. The powder of Claim 31 wherein the macroparticles are composed principally of silica and wherein the component of the bimodal pore size distribution having the smaller average pore size provides about 40 to about 55% of the total pore volume and the component of the bimodal pore size distribution having the larger average pore size provides about 60 to 45% of the total pore volume.

33. The powder of Claim 32 wherein each component of the bimodal pore size distribution provides about 45 to 55% of the total pore volume.

34. A powder of Claim 27 wherein said macroparticles are totally porous, each being composed of a plurality of microparticles, the microparticles having a diameter of about 0.005 to 1.0 micron.

35. A powder of Claim 33 wherein the macroparticles have a diameter of from about 5 to 50 microns and the microparticles have a diameter of 0.005 to 0.5 micron.

36. The powder of Claim 35 wherein the macroparticles are composed principally of silica and wherein the component of the bimodal pore size distribution having the smaller average pore size provides about 40 to about 55% of the total pore volume and the component of the bimodal pore size distribution having the larger average pore size provides about 60 to 45% of the total pore volume.

37. The powder of Claim 36 wherein each component of the bimodal pore size distribution provides about 45 to 55% of the total pore volume.

38. The powder of Claim 27 wherein the macroparticles are totally porous having a core of a plurality of ultramicroparticles about 1 to 30 nanometers in diameter and a skin of a plurality of microparticles about 0.1 to 1.0 micron in diameter.

39. The powder of Claim 38 wherein the macroparticles are 5 to 50 microns in diameter and the microparticles are 0.1 to 0.5 micron in diameter.

40. The powder of Claim 39 wherein the macroparticles are silica and wherein the component of the bimodal distribution having the smaller average pore size provides about 40 to 55% of the total pore volume and the component of the bimodal distribution having the larger pore size provides about 60 to 45%
of the total pore volume.

41. The powder of Claim 40 wherein each component of the bimodal pore size distribution provides about 45 to 55% of the total pore volume.