WO1994006535A1

WO1994006535A1 - Apparatus and method for fractionating a liquid mixture

Info

Publication number: WO1994006535A1
Application number: PCT/US1993/008523
Authority: WO
Inventors: Halbert Fischel; Richard J. Fischel
Original assignee: Halbert Fischel; Fischel Richard J
Priority date: 1992-09-11
Filing date: 1993-09-09
Publication date: 1994-03-31
Also published as: AU5126693A

Abstract

Apparatus and method for fractionating a liquid mixture into a plurality of liquid fractions whose components have different velocities. A rotating cylindrical assembly (A) has inner and outer shells (I, O) defining therebetween a tubular gap (G). One shell includes a porous structure (21) for permeation therethrough of one or more of the liquid fractions and to resist radial inward flow of the remainder. The cylindrical assembly (A) is rotated at a speed effective to impart sufficient centrifugal force to the liquid fractions to maintain one such fraction within the gap (G) while compelling flow of the other liquid fraction radially inwardly through the porous structure. The invention includes a porous structure (21) in both the inner and outer shells (I, O) to allow for levitation of one fraction of the mixture between such structures while a wash of solution bearing beneficial components is flowed radially inwardly therethrough.

Description

APPARATUS AND METHOD FOR FRACTIONATING A LIQUID MIXTURE

FIELD OF THE INVENTION

The present invention relates to centrifugation apparatus and methods for separating liquid mixture components, as for example, blood, having different sedimentation rates.

BACKGROUND OF THE INVENTION

Prior centrifugal separation systems, either continuous or batch, generally form layers of substantial thickness. The spin axis defines the rotation that produces the centrifugal force, which force causes the eventual separation of components according to differing densities. Greater densities occupy comparatively greater depths when sufficient time is allowed to permit the effective completion of the sedimentation process. The depth of a layer refers to its distance from the spin axis. As some components have a distribution of densities which overlap the density distribution of other components, separation of such components by this means is necessarily incomplete. Even with no overlap of component density distributions, separation may not be complete or as perfect as desired due to instabilities in the suspending media under rotation making it particularly difficult to extract material from thin intermediate layers. Such instabilities tend to increase as total liquid thickness increases, especially when the material is caused to flow for the purpose of decanting various components.

Parameters, such as the time required to produce useful separation by density layering, rotation rates, layer thickness, decanting geometry, volume processing capacity, and component purity have, heretofore, required undesirable trade-offs in which one was compelled to compromise purity as against separation time or volume processing capacity. In most cases, limitations are imposed by the decanting geometry, method and rate which may introduce disturbances in the formed layers especially when the latter are very thin as in the case of "white" cells of blood, i.e., the "buffy coat".

Regardless of the many and diverse techniques utilized by prior art centrifugal separators for dealing with the various problems just described, they all require the separated material to form layers which closely approximate the end point of the sedimentation process. The loci of separation as between the several layers can be used to define a surface normal vector that would be substantially perpendicular to the spin axis at all points of the layer. Generally, the decantation of any particular layer requires at least some component of flow transverse to the layer's "surface normal" vector. See, for example, U.S. Patent No. 4,086,924. It is easily seen that one layer is required, at least in part, to "flow" in a shearing fashion over other layers in order to be decanted. This necessarily introduces stress on the interface between layers, which can contribute to contamination between layers. Viscous shear flow between fluid layers is anathema to centrifugal separation by density selective layering due to the instabilities it introduces under a rotational force field. Nevertheless, at least one component of shear flow along separation interfaces has always been involved at some stage of separation in prior centrifugal separators.

Other prior art techniques deliberately utilize shearing of suspensions flowing over membranes to separate components by what is understood to be cross- flow filtration (see A.L. Zydney and C.K. Colton, Continuous Flow Membrane Plasmapheresis: Theoretical Models for Flux and Hemolvsis Prediction.. Trans ASAIO, Vol 28 (1982), p. 408.). However, membranes become clogged over time.

Efforts to prevent clogging of the membrane have included increasing the membrane area (see B.A. Solomon, et al, U.S. Patent No. 4,212,742), the shear rate of the flowing blood so as to delay the clogging effect (see R.J. Fischel, et al, U.S. Patent No. 4,755,300), or the pressure differential across the membrane (see F. Castino, et al. The Filtration of Plasma from whole Blood: A Novel Approach to Clinical Detoxification. In: Chang, T.M., et al., Artificial Kidnev. Artificial Liver, and Artificial Cells. New York: Plenum Press, 1978, pp. 259-266) .

Another technique that has long been used for cell separation is usually referred to as centrifugal elutriation or CE (see M.L. Meistrich. Experimental Factors Involved in Separation by Centrifugal Elutriation. Cell Separation Methods and Selected Applications, Ed: Pretlow, T.G., Academic Press, New York, V.2, PP: 33-61 (1983)). The basic principle involved in various embodiments of this technique is separation by differences in particle settling rates. In some instances, a counter-flowing fluid is introduced against the centrifugal force field (see R.L. Berkow et al Purification and Functional Evaluation of Human Polymorphonuclear Leukocytes. Cell Separation: Methods and Selected Applications, Ed: Pretlow, T.G., Academic Press, New York, V.4, PP: 147-170 (1987)), and is referred to as counterflow CE or CCE. These methods are intended as experimental sample batch processes. They are not suitable for rapid processing of large volumes on a continuous basis. The blood cell washing procedure of the present invention uses much less costly sterile washing solution than required by prior apparatus employed for this purpose. The speed and thoroughness of blood component separation without the unacceptable blood priming volumes or harsh treatment of the fragile blood elements, not otherwise achieved heretofore, is now possible with particular embodiments of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic central vertical section of a apparatus embodying the present invention showing the basic method of operation thereof;

FIG. IA is broken, fragmentary, schematic and enlarged view of the right-hand portion of FIG. 1 demonstrating the behavior of fluid components relative to the porous element;FIG. 2 is a horizontal sectional view taken along line 2-2 of FIG 1;

FIG. 3 is a broken foreshortened central, vertical sectional view showing the major components of the form of apparatus of FIG. 1 embodying the present invention;

FIGS. 4 and 5, are horizontal sectional views taken along lines 4-4, and 5-5, of FIG.3; FIG. 6 is a foreshortened central vertical sectional view showing major components of a second form of apparatus embodying the present invention;

FIG 6A is a fragmentary central vertical sectional view of the apparatus of FIG. 6 showing the parts thereof arranged for a washing function;

FIG. 7 is a horizontal sectional view taken along line 7-7 of FIG. 6. FIG. 8 is a central vertical sectional view showing major components of a third form of apparatus embodying the present invention.

FIG. 9 is a horizontal sectional view taken along line 9-9 of FIG. 8.

FIG. 10 is a graph demonstrating one principle of the invention relating to the separation of the components into separate fractions;

FIGS. 11 and 12 are schematic views demonstrating the relationship between rotational speed of the centrifugation unit and movement of the liquid components during operation of the apparatus embodying the present invention;

FIGS. 13, 14 and 15 are plots of diameter, density and settling rate, respectively, of various blood components in terms of population distribution, with FIG. 15 referring to Stokes settling velocity per unit of centrifugal force;

FIG. 16 is a schematic illustration showing how two devices embodying the present invention can be operated as two simultaneous stages for collection of platelet concentrate;

FIGS. 17-20 are schematic illustrations showing various functions which may be obtained employing the method and apparatus of the present invention.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

There are several important principles applicable to the design and operation of the method and apparatus of the present invention. Critical to the preservation of stability of centrifugal settling is the concept of spinning a "thin", as opposed to a "thick", film. The parameters which distinguish the two will be computed for blood and presented hereinafter. Initially, it is important to understand that continuous operation of the method and apparatus requires that blood, or other liquid mixture, enter the spinning "film" at one end, preferably the top, and move axially, downwardly in laminar stream flow, i.e., viscous boundary layer flow, approximately parallel to the spin axis toward an exit from the "film". This flow would not be stable, in the sense that various types of secondary flows would be triggered, thus upsetting the preferred radial convective flow and particle settling geometry, if the film were not sufficiently thin. Thicknesses available to blood under this restriction lead to practical and useful designs.

Referring now to FIGS. 1 and 2, a preferred method and apparatus for separating a liquid mixture, e.g., blood plasma, and a blood component or components in that liquid with the liquid and blood component or components forming a flowable mixture is schematically illustrated. The apparatus includes a cylindrical assembly A adapted for rotation about a central vertical axis 20 and having inner and outer generally cylindrical shells I and O, respectively, defining therebetween an axially extending tubular gap G. The inner shell includes a porous cylindrical structure 21 for at least a part of its axial extension for a purpose to be described hereinafter. The cylindrical assembly A is shown disposed within a fixed housing H, for rotation relative to the housing about axis 20. The flowable liquid mixture which is to be fractionated into first and second liquid fractions, containing components having settling velocities which are higher and lower respectively than a selected value, enters the upper end of the gap G between the inner and outer shells. Fractionation is effected within this gap between spaced apart generally axially concentric and rotationally symmetric surfaces 32 and 34 of the inner and outer shells, such surfaces forming boundaries as defined by walls of the inner and outer shells. The boundaries 32 and 34 are shown as concentric tubular boundaries extending about the axis of rotation 20 of the cylindrical assembly A, and may either be cylindrical or variously, somewhat differently tapered, i.e., inclined to the rotational axis of the assembly at particular axial locations, as described hereinafter. The direction of rotation of the cylindrical assembly is indicated by arrow 36, and the liquid mixture, or fractions, in various states of localized concentration of particular components flow generally downwardly, i.e., axially from the top of gap G to the bottom of gap G between the vertically orientated boundary surfaces 32 and 34.

A vertical support tube 41 rigidly interconnects and carries the inner and outer shells I and 0 at 42 and 44. Power operated means (not shown in FIGS. 1 and 2) are interposed between housing H and the cylindrical assembly A to spin the inner and outer shells and their confronting surfaces 32 and 34 about the common axis of rotation 20 at the same angular velocity. The liquid mixture to be fractionated M may be fed into the axial portion of gap G above an imaginary reference horizontal plane 81 and also pressurized by means not shown in FIGS. 1 and 2 , as via a feed stream indicated by flow path 46. Liquid mixture components in the gap more dense than the liquid mixture continuous phase migrate radially outwardly during rotation of the cylindrical assembly due to centrifugal force which is exerted on all elements of the liquid mixture whereby concentration of the said more dense components in the liquid flowing axially (shown) downwardly in FIG. 1 in the gap G becomes greater and less near surfaces 32 and 34, respectively. If the process of liquid component migration were allowed to proceed to completion in the gap, a boundary marking the separation of the liquid continuous phase from the settled discontinuous liquid components could, in some cases, become visibly well defined and is depicted by an imaginary reference boundary 48 which need not actually form as shown in the drawings. Boundary 48 is an imaginary cylindrical surface essentially parallel to the vertical axis of rotation 20 having a radius which can be controlled by mass balance of the liquid mixture and fraction flows into and out of the apparatus as described more fully hereinafter.

Controlled radially inward flow from gap zone 49, i.e., the axial portion of gap G below horizontal plane 81 of the second liquid fraction consisting of at least the continuous liquid phase of mixture M or a continuous phase including one or more discontinuous components in that zone is enabled in a pressure-drop direction operating to compel flow out of the gap zone 49 and through the porous surface 34 and the porous element 21 of the inner shell I into an inner region R defined between the inner surface of inner shell I and a cylindrical barrel element 50 coaxial with tube 41. Such enable ent is effected as a function of the static pressure difference between gap G and interior region R, the effect of angular velocity of rotation ω of the zones defining boundaries, subjecting the liquid to centrifugation pressure opposing inward radial flow, and the porosity of porous structure 21 which resists flow. The thickness of the gap G will vary depending upon the nature of the liquid mixture being fractionated. In the case of blood, the preferred gap thickness is between about 5 to 150 mm. As noted, the fluid mixture progresses axially and downwardly as it encounters the porous structure 21 of the inner shell, the primary purpose of which is to act as a flow distributor. In order to serve this function, the porous structure must have certain properties. First, of course, the pores must be large enough to freely pass without sensible restriction, the discontinuous or formed constituents to be decanted. The pores may offer no restriction to constituents or particles that are intended to remain behind in the gap with the original medium because "filtration", understood as a sieving effect, plays no part in the operation of the method and apparatus of the present invention.

On the other hand, the porous structure must offer flow resistance to that portion of the suspending medium carrying the decanted constituent in the sense of creating a "pressure gradient". The preferred operative requirement is that the Trans Porous Pressure (TPP) drop substantially exceed the pressure loss required to move fluid from its entrance to the spinning film (mixture M) to its exit from the film (first liquid fraction M.,) . Consequently, the porous structure means can act as a largely "uniform" flow distributor in accordance with the invention. In practice, this is not a serious restriction in, for example, generally useful blood flow rates, given that the thin film flow cross-sections possible in accordance with the invention, do not lead to significant axial blood flow pressure losses. Furthermore, as discussed in more detail within, there are very effective ways to increase the flow resistance of the porous means so that TPP is always sufficiently greater than the axial pressure gradient to cause substantial uniformity of decanting, i.e., uniform convective radial velocity over the entire porous surface. The uniform decanting velocity referred to here is defined by dividing the volumetric decanting rate, of second liquid fraction, M₂ (ml/sec) by the porous area (cm²) . The uniform decanting rate is adjusted by controlling input and output volumetric pumping rates of incompressible fluids.

Centrifugation of the axially (here downwardly) flowing liquid mixtures in gap G tends to move discontinuous components throughout the mixture and separating fractions, which are more dense than the suspending continuous medium, in an outward radial direction generally opposite the inward radial flux through porous surface 34 and away from axis 20. Components which are less dense than the suspending continuous medium (i.e., buoyant particles) move in an inward radial direction under the influence of centrifugal force and are carried with the inward radial flux in all cases. Absent inward radial flux, the velocity of movement of discontinuous components in the radial direction due to centrifugation characterizes their effective settling rate.

Velocities vary among the individual specie of components and tend to retard as concentration of all components taken together increases. In dilute concentration the settling velocity, V_s, of a particular species of component, tends to be narrowly distributed and predicted by the Stokes equation. It is effectively proportional to centrifugal force, density difference, p over that of the suspending medium and the square of particle diameter, D, (or dimension characterizing particle size) and inversely proportional to suspending medium viscosity, μ . Centrifugal force is proportional to the distance from the rotation axis 20 and the second power of the angular velocity, ω. Consequently, centrifugation may be controlled, as for example by control of angular velocity, given radial dimensions of boundaries 32 and 34, to allow permeation of any component or components having radially outward characteristic settling velocities below a selected value, in the inward radial direction, along with the liquid, i.e., controlled permeation through the porous boundary 34 and porous cylindrical structure 21, the porosity of the latter in general allowing such components in the mix to pass through unobstructed. In one example, the porosity may be such as to allow all components (of different sizes and settling rates) to pass, i.e., including those of a relatively greater settling rate, but which are not allowed to permeate, due to the angular rate ω of centrifugation tending to retain them in the gap zone 49.

Flow path 54 in FIG. 1 indicates removal from gap G of modified mixture M,, in the form of, or consisting of, the initial liquid mixture M less any liquid which has flowed radially inwardly through porous structure 21, and less any particle component or components carried with the liquid permeating porous structure 21. In particular, the difference between the rate or amount of feed of mixture M into the gap along path 54 at the top and the rate or amount of removal of M₁ from the gap along path 54 at the bottom is necessarily the net rate or amount of permeation, i.e., decantation for an incompressible liquid such as blood along flow path 46. In practice, static pressure is actually controlled by controlling the feed flow and removal flow rates using positive displacement pumps (not shown in FIGS. 1 and 2). FIGS. 11 and 12 illustrate vectorally the forces on the particles due to centrifugation and decantation. With continued reference to FIG. 1, it is important to note, that the upper portion of the inner shell is formed with a impervious inverted cup-shaped entrance member 66, the sides 68 of which are tapered downwardly and inwardly away from boundary 48 towards the center of the cylindrical assembly A. The lower end of entrance member 66 abuts the top of porous element 21 at horizontal plane 81.

Outer surface, 34 of porous element 21 forming the interior surface of the gap G is either parallel to or preferably tapered outward from axis 20 relative to boundary 48. The outer cylinder 0 also preferably tapers outward from axis 20 at an angle usually similar to that of surface 34, for reasons more fully explained below. Surface 68 is seen to resemble an inverted frusto-conical surface while surfaces 32 and 34 are seen to resemble upright frusto-conical surfaces, respectively. The angle of taper and special shaping, if any, of surfaces 68, 32 and 34 may best be understood by treating boundary 48 as a downward axially moving surface which matches the velocities at boundary 48 associated with the axial flow of liquid in gap regions 82 and 83, respectively. Gap regions 82 and 83 are radially inward and outward of cylindrical boundary 48, respectively. The purpose is to minimize shear, i.e., axial velocity gradient, as between regions 82 and 83 at boundary 48. By limiting shear between fluids of substantially different viscosities in a rotational field subject to strong Coriolis forces, laminar stream viscous boundary flow can be preserved as more fully explained within.

As less highly concentrated fluid (i.e., much of the suspending medium of the original mixture M) is drawn away from that part of region 82 below horizontal plane 81, the axial flow cross-section of that region must be reduced in order to maintain matching axial velocity between fluids in regions 82 and 83, respectively at boundary 48. As more concentrated fluid forms in regions 83 below horizontal plane 81, the opposite must obtain. Consequently, solid surface 32, and porous surface 34 have the appearance of frusto-conical sections tapering (downward) away and toward cylindrical boundary 48, respectively. The outward taper of surface 32 further assists the downward (outward) axial flow of concentrated material by providing a resolved component of centrifugal force tangent to surface 32 in an axial direction toward the exit region 40 for the first fraction. The portion of total centrifugal force assignable to the outward axial component is the sine of the angle surface 32 makes with rotational axis 20. Even at one part in one hundred the axial component of force which must be added to one gravity in vertically oriented apparatus can be considerable and always advantageous to the movement of thickened viscous material.

The radial position of boundary 48, to the extent it actually forms, is best understood in relation to the lower extremity of surface 34. If most of the continuous phase contained in the original mixture M iε decanted as suspending medium for the second fraction M2 or is itself M2 then boundary 48 will closely approach the lowest extremity of interior surface 34 because very little axial flow cross- section is required for the remainder of M2 to enter zone 62 where it re-mixes with the concentrated material leaving the lower end of gap G to become the final formulation of Ml. In the limit where all available M2 is decanted and concentrated material is well formed, that is, boundary 48 is well defined, the latter will just intersect the extreme lower corner of surface 34. Even more aggressive decantation of M2 will cause some of the concentrated material to enter M2 and boundary 48 intersects surface 34 at some axial position above the latter's extreme lower corner. Thus, the radial position of a well formed boundary 48 is very sensitive to the fraction of _M2 decanted through porous surface 34. Boundary 48 is merely a convenient reference for purposes of illustrating the process, however, the permeation of what is, at first, a very minuscule amount of concentrated material into M2 can be used as an observable indicator of conditions within the apparatus in most cases. Nephelometry or hemoglobin detection means placed at the output 86 of M2 will detect the very first permeation of cells or RBC, respectively, in the case of blood, whereby decantation flows may be adjusted at will. Similar techniques are available for most mixtures having concentrated separable components.

The above described liquid fractionating phenomena can be more fully understood by reference to FIG. IA, when taken in conjunction with FIG. 1.

Referring now to FIG. IA, there is shown a broken, fragmentary schematic enlarged view of the right-hand portion of the cylindrical assembly A as it is rotated, depicting the flow pattern of the first and second liquid components Ml and M2, respectively, relative to the porous cylindrical structure 21. For purposes of illustration, the space within gap G has been divided into zones RA, RB, RC, and RD. Zone RA is the space radially inward of boundary 48 radially outwardly of the cup-shaped entrance member 66. Zone RB is the space radially outwardly of boundary 48 adjacent zone RA toward which all of the mixture components (more dense than the suspending liquid medium and thus subject to centrifugal force) move. Zone RC is disposed below the imaginary horizontal plane 81 adjacent the outer surface 34 of porous element 21 within boundary 48. Zone RD is disposed adjacent zone RC, but radially outwardly thereof and the porous element surface 34. Within zones RA and RB, all of the said more dense components that are subject to centrifugal force move radially outwardly, and there is no opportunity for radially inward flow. The purpose of including these zones is to provide an opportunity for the concentration of otherwise concentrated particles nearest the spinning outer surface of the cup-shaped element 66 to be lowered so as to create dilution of such particles radially inwardly of boundary 48. Within zone RB, increasing concentration of particles is taking place which tends to make boundary 48 more visible. Such boundary could, in fact, become readily visible when the liquid mixture is blood, since the red blood cells, as they start collecting within zone RB become quite dense. As plasma thins out, it begins to appear yellow and clear, even though plasma may be full of platelet particles. Zone RC is the zone which is most proximate to the outer surface 34 of the porous element wherein the separation of component M2 iε occurring. Within zone RC, a predictable or characteristic radially outward settling of particles takes place, since zone RC now contains a diluted mixture of particles. Because a known radially inward flow of liquid, possibly including certain particles and a known outward radial settling velocity of all particles is established in dilute concentration, it is possible to decide whether or not particular particles will be carried with the liquid leaving zone RC through the porous element 21. In zone RD, the particles are becoming more heavily concentrated due to centrifugal force, and accordingly, will not flow radially inwardly through the porous cylindrical element 21. The fact that the liquid within zone RA is becoming more dilute means that it is becoming less viscous. The amount of dilute liquid is increasing in volume as the mixture flows axially downward so that the liquid in zone RA is increasing in volume.

Within zone RB, the liquid is becoming more concentrated, and accordingly, becomes more viscous. In zone RC, the liquid is less viscous, as in zone RA, but is also being removed radially inwardly through the porous element, and is accordingly, losing volume as the volume increases in zone RA. In zone RD, the liquid is becoming thicker and is substantially increased in viscosity so as to require more space within which to flow. The aforedescribed tapering of the outer and inner facing surfaces of the outer and inner cylinders is designed to take into account the fundamental physical fact that the division between the liquid portion and the concentrated portion follows a surface that is essentially parallel to the axis of rotation 20, and it should be understood that such cylinder is indicated by boundary 48. Boundary 48 is located radially within the gap depending upon the volumes that are flowing in and out of the cylindrical assembly. Because of the axial downwardly flow of the liquid, in combination with the radially inward flow of liquid component M2, the axial flow and the radial flow are vectorially independent as noted hereinabove. The aforedescribed tapered surfaces insures approximately the same axial velocity for the concentrated and non-concentrated liquid components at boundary 48. If such approximately matched axial velocity is not obtained, there would be a tendency for one or the other liquid component to have a shearing effect that would tend to create vortices in the high energy rotational field generated by rotation of the cylindrical assembly A.

One of the most important criteria for a device dedicated to the collection of blood components is the ability to isolate all surfaces and environments that could be exposed or accessible to the blood or its separated products. These must be sterilizable by means approved in the industry (e.g., Gamma, ETO, etc.) and must be capable of retaining sterility under ordinary conditions of use. This is extremely difficult and expensive to achieve when attempting to seal a moving part from the external environment. The solution of the problem in connection with the present invention as disclosed herein is not, in principal, restricted to the present invention, rather it is believed it constitutes a significant contribution to the practical embodiment of devices incorporating, moving, sterile parts.

The basic concept provided by the present invention is to immerse the entire rotating cylindrical assembly A in a bath of sterile solution, compatible with blood, e.g., saline, or anticoagulant, indicated by flow arrows 90, which is contained within the stationary housing H. Access to the interior of the housing H is made through upper and lower conduits 91 and 92, to be described fully hereinafter, which lead into and out of the housing from or to sterile containers respectively. The entire sterile path, including the rotating cylindrical assembly, and the several sterile containers which may be permanently attached to the assembly housing prior to sterilization comprise a closed sterile system. Conventional peristaltic pumps, not shown in FIG. 1, engaging conduits in communication with the interior of the assembly can be used to move fluids to and from the containers and the assembly without invading the sterile environment which includes, in part, the extracorporeal blood path and the vessels in which the separated concentrated blood components are stored. See FIG. 18. As indicated by flow path 90 which begins at the upper portion of the rotating assembly and its stationary housing H, a sterile or wash solution is urged downwardly between the space 93 separating the exterior surface of the outer cylinder O and the inner surface of the stationary housing H. Such sterile or wash solution is discharged at the bottom of the rotating assembly and housing combination. Fluid seals (not shown in FIGS. 1 and 2) are interposed between the upper and lower elements of the rotating assembly and its stationary housing H. A portion of the sterile solution flows through these seals.

It is important that the sterile bath is maintained at a higher static pressure than the fluid flowing through paths within the rotating assembly. With this arrangement, the entire rotating assembly is surrounded by the sterile bath and the seals which need not function as bearings, are provided to operate between that bath and the fluid flow paths of the rotating assembly. The criteria governing the design of the seals (preferably rotating seals) are less critical than the criteria normally required for sterility barriers. For example, although these rotating seals (as between the sterile bath and the blood) can be impervious to fluids without detracting from the operation of the apparatus and method of the present invention, because the sterile bath is maintained at a higher static pressure than the blood, or, in the alternative view, caused to flow at a pre¬ determined rate across the seal by means of positive displacement into the blood, as long as the leakage rate is small, the dissolution effect is harmless, and the direction of flow operates to protect the fragile components of the mixture by preventing them from entering the seals. As described hereinafter, however, as in cell washing, the sterile bath may also be the wash solution and a high leakage rate is intended.

The advantages of such a rotating seal concept are significant in connection with high spin rates. First, high sealing and bearing pressure is not required, which mitigates the heating effect. Second, the sterile fluid crossing the sealing surfaces at their interface forms a fluid film that reduces friction, further mitigating the heating effect while absorbing and carrying away what heat is generated. The same concept may be employed in separate load bearing bearings between the housing H and cylindrical assembly A, except that flow through the bearings is much higher and does not represent a "leak" from one fluid into another. The cooling effect of the sterile bath can be enhanced by maintaining a circulation of this bath around the rotating mechanism and through the rotating bearings and seals exposed to the circulating bath. The same sterile bath can be continuously recirculated by providing fluid entrance and exit routes on the stationary containment vessel, as shown in FIG. 1 for that purpose.

It should be noted that not maintaining a positive flow across the rotating seal from sterile bath into the blood path does not result in a critical difficulty. There is no loss of sterility and, contamination of the sterile bath with blood or blood components is not crucial since this material is disposed of, following the procedure. Nevertheless, it is not desirable because blood components, particularly RBC, will inevitably be damaged in transit across the nominal seal. The damaged components carried into the sterile bath are of no particular importance, but the material they generate (e.g., free hemoglobin) can and usually does leak back into the primary blood path. This is to be avoided in the manner previously described in keeping with the highest quality blood separation. Heretofore, systems using rotating seals against blood without a protective leakage component, as described herein, have invariably caused measurable contamination with hemoglobin released by RBC damaged within the seals.

Referring now to FIGS. 3, 4, and 5, there is shown the major components of a first form of apparatus embodying the present invention constructed in accordance with the principles of the invention described hereinabove and shown in FIGS. 1, IA, and 2. Like parts bear the same reference numerals. The power-operated means for rotating the cylindrical assembly A within housing H may take the form of a conventional electro-magnetic armature drive utilizing a stator 100 affixed to the upper part of housing H which cooperates with an armature assembly 102 which is keyed to an upper coaxial drive tube 104 that extends downwardly with its lower end affixed to the upper portion of the support tube 41. An upper bearing member 115 is interposed between the upper portion of drive tube 104 and a drive support member 105 that extends upwardly from a neck 107 formed on housing H. The upper end of drive support member 106 is provided with an inlet fitting 110 which is connected with the upper end of drive tube 104 to conduct the liquid M downwardly therethrough. An upper combination bearing and O-ring assembly 112 is interposed between the upper portion of drive tube 104 and the drive support member 106. A sterile solution conduit 91 is affixed to the neck 107 of the stationary housing H to conduct sterile solution upwardly through a first sleeve 113 formed on the neck along the inner surface thereof, radially inwardly and downwardly through vertical passages 114 of upper bearing 115, then radially outwardly below bearing 115 and downwardly along the inner surface of a second sleeve 116 below drive support member 105 and exterior of armature assembly 102. From the space below the armature assembly 102, the sterile solution flows downwardly through a main bearing assembly 118 through vertical passages 119 formed through the intermediate bearing assembly. Main bearing 118 is carried by neck 107.

Thereafter, the sterile solution flows downwardly into space 93 between the interior of the housing H and the exterior of the outer shell O. Next, the sterile solution flows downwardly through the lower confines of the space 93 underneath the outer shell O to flow downwardly out of a discharge conduit 92. It should be noted that additional sterile solution is introduced into the base portion 120 of the housing H through a conduit 130, such sterile solution being indicated by flow arrows 132. The sterile solution entering through conduit 130 flows upwardly through vertical passages 133 formed in a lower bearing assembly 142 to mix with the sterile solution entering through the upper conduit 91, and then flows outwardly through discharge conduit 92. This second flow of sterile solution also moves downwardly through a lower top combination bearing and O-ring seal assembly 140 and upwardly through a lower combination bearing and O-ring seal assembly 142. The sterile solution may serve as a coolant for the bearings and seal assemblies.

Referring to the upper portion of FIG. 3, the liquid component M designated by flow arrows 54 is moved downwardly through drive tube 104 and radially oriented in passages 141 formed in the common top wall of the inner and outer cylinder, then into the upper portion of gap G to flow downwardly therethrough to be separated as described herebefore into the first liquid component Ml. As indicated by flow arrows 144 liquid Ml flows radially inwardly through passages 145 formed in the common bottom wall of the inner and outer shells, and then downwardly through lower tube 150 to be finally discharged through discharge conduit 86 which is coaxial with the tubes and the cylindrical assembly. It will be seen that the lower tube 150 extends through bearing O-ring seal assemblies 140 and 142 and a bottom combination bearing O-ring assembly 154. The sterile solution entering through lower conduit 130 cools the lower bearing and combination bearing and O-ring seal assemblies.

With continued reference to FIG. 3, liquid component M2 which flows through the porous cylinder 21, as indicated by flow arrows 46, moves radially inwardly from the bottom of inner region R through passages 160 formed in the lower portion of the inner and outer cylinders I and O to be collected in chamber 164 formed between the lower portion of the vertical support tube 41 and the lower drive tube 150 for ultimate discharge through fitting 87.

Shock absorbing pads 156 may be interposed between the horizontally facing surfaces of housing H and outer cylinder 0.

It should be understood that the sterile solution entering through the respective upper and lower conduits 91 and 130 follows a sterile path which includes the exterior of the rotating cylindrical assembly A and is maintained at a higher static pressure than the fluids M, Ml, and M2 so as to protect the fragile components of these fluids by preventing them from entering the seals as described in detail hereinbefore. It should also be understood that it is possible to force a saline solution into the gap G, as hereinbefore described, should it be necessary to re-liquify or otherwise render fluid any packed cells of the liquid component Ml within the lower portion of gap G.

The principles of operation of the separator apparatus shown in FIG. 3 may be seen by reference to the schematic illustration of FIG. 18. Referring thereto, the housing A4 of the separator apparatus receives blood from a donor at its upper end. Upon rotation of the cylindrical assembly, the concentrated blood cells of the donor are returned to the donor, as indicated by the directional arrows. Plasma is transferred to a suitable receptacle from the lower right side of the housing as shown by the directional arrows. The volumetric pumping rate of the blood- handling upper pump Pll is greater than that of the lower pump P12 so as to provide the pressure differential required to effect permeation of the plasma through the walls of the porous cylinder element 21. Saline sterile solution forced by pump P13, enters the upper and lower left-hand portions of the housing A4 and such saline solution is metered out of the lower portion of the housing by pump P14 as shown by the directional arrows. Saline pumps P13 and P14 maintain the saline solution at a higher pressure than the pressure maintained in the separator apparatus and the conduits leading to and from such separator for the reasons set forth hereinbefore.

Referring now to FIGS. 6, 6A, and 7, there is shown the major components of a second form of apparatus embodying the present invention. This form of the invention is similar to the form of the invention shown in FIGS. 3, 4, and 5, with the exception that the inner shell I' is formed at its lower portion with a second porous structure in the form of a porous ring 170 disposed below the main porous cylinder 21'. It should be understood that the upper part of the apparatus of this second form of the invention is also provided with a electro-magnetic armature drive (not shown) disposed within a drive support member 106' of the type utilized in the apparatus of FIG. 3 for rotating the cylindrical assembly A as a liquid M to be fractionated is caused to flow downwardly within drive tube 104'. During such rotation, component migration, particularly described in conjunction with the description of FIG. 1 and IA hereinbefore takes place whereby an imaginary boundary 48' is formed in general alignment with the upper and lower peripheries of the inner shell I' as the liquid fraction M2 flows radially inwardly through the main porous element 21'.

In this form of the apparatus, however, a third liquid fraction M3 is harvested from the lower portion of the fluid within the radially inner lower zone E portion of the boundary 48'. Such liquid component fraction M3 is forced through the lower porous ring 170 by the excess of fluid pressure within the gap G over that maintained in an inner region RR interior of porous ring 170. Such liquid, after passing through the lower porous ring 170 continues to move radially inwardly through the lower region RR which is defined by passages 174 formed in the lower walls of the inner and outer shells. This M3 liquid fraction is collected within a space 176 disposed outwardly of lower tube 150' to be ultimately discharged through an outlet conduit 178 formed in the lower right-hand portion of the stationary housing base 120'.

With continued reference to FIG. 6, 6A, and 7, the lower portion of the outer shell 0', generally radially outwardly of the porous ring 170, is formed with a cylindrical opening 180 over which is disposed a vertically movable cylindrical valve ring 182. Such valve ring 182 is seated on a pair of O-ring seals 183 mounted in the outer portion of opening 180. The lower portion of the valve ring is formed with a plurality of circumferentially spaced holes 186. A porous diffusion ring 184 is disposed within the reduced diameter opening 185 inwardly of and aligned with opening 180. Valve ring is normally arranged in its lowered position of FIG. 6, however, such valve ring may be moved upwardly into its raised position of FIG. 6A by suitable electro-magnetic means (not shown) in a conventional manner. Steps 188 limit movement of the valve ring. In the raised positions of the valve ring 182, the holes 186 of the valve ring are disposed radially outwardly of the openings 180 and porous diffusion ring 184. With the valve ring is arranged in its normal lowermost position, saline solution entering through upper and lower conduits 91' and 130' flows outwardly through an outlet conduit 190 attached to the left-hand lower portion of housing H' from the space between the outer surface of the outer shell O' and the inner surface of stationary housing H'. The saline wash solution entering lower conduit 130', additionally, flows through the lower bearing and bearing O-ring seal members, just as in the apparatus of FIG. 3 and for the same purposes. When the valve ring 182 is moved to its raised position of FIG. 6A, however, saline wash solution is introduced radially inwardly through valve ring holes 186, as indicated by flow arrows 194, whereby such saline solution will flow radially inwardly from space 93' through the porous rings 184 and then ring 170 at a rate of flow effective to enable an inward radial velocity of controllable and uniform value within the gap G'. Diffusion ring 184 may be somewhat displaced axially upward (i.e., toward the entrance zone of the gap) relative to porous ring 170 in order to accommodate downward axial displacement of wash solution in the time required for the latter to cross the gap. After the saline wash solution enters and crosses porous outer ring 184 and gap G', it enters and crosses porous ring 170, and enters inner region R' , as noted above. The inward radial convective velocity associated with wash solution flow should be higher than that of prior flow across surface 34' of porous element 21' in order to take full advantage of a sequential concentrating followed by a washing step intended to remove contaminants not otherwise permitted to follow with flow across surface 34'. Forced inward convection of wash solution into and across the concentrated cell mass tends to displace the cellular material radially inward away from the interior surface of porous outer ring 184, i.e., floated or "levitated" off the surface. Each cell that remains with the first fraction finds a new point of equilibrium (at some radial position within the gap) where convective drag on the cell (a force which is controlled by wash solution flux) is in balance with centrifugal force on the cell. Any particle which cannot find new equilibrium, e.g., broken cell stroma, will be washed out. It should be understood that the portion of boundary 48' between the facing porous surfaces of rings 170 and 184 will become somewhat displaced radially inward from its position as shown in FIG. 6A and less clearly defined because the population of cells has a distribution of settling velocities. Some perturbation of reference boundary 48' must occur in zone E due to the change in equilibrium conditions previously established in zones RC and RD'. Liquid component M2 which flows through the porous cylinder 21', as indicated by flow arrows 46', moves radially inwardly into inner region R' and then upwardly through such region and into passages 199 formed in the upper common walls 198 (Fig. 6) of the inner and outer shells and then upwardly through a space 201 between the outer surface of drive tube 104' and the inner surface of support tube 41' into a collection chamber 202 between combination bearing and O-ring seals 203 and 204 to be discharged through a conduit 205. It should be understood that during operation of this form of the apparatus, sterile solution is forced through conduit 91' into the housing neck 107' so as to cool the bearings and seals disposed in the drive support member 106' and the housing neck in the same manner described hereinbefore with respect to the apparatus of FIG. 3. It should also be understood that the sterile solution entering through conduits 91' and 130' follows a sterile path which includes the interior of the housing H' and the exterior of the rotating cylindrical assembly A, and is maintained at a higher static pressure than the fluids M, Ml, M2, and M3 so as to protect the fragile components of these fluids by preventing them from entering the seals as described in detail hereinbefore. Moreover, the introduction of sterile solution through porous element 184 makes it possible to re-liquify or otherwise render axially flowable any packed cells of the liquid component Ml within the lower portion of gap G'.

Reference may now be made to FIGS. 13 - 15 for a description of the general mode of operation of the present invention as applied to previously described separator apparatus. It is important to note: a) the separation of particles is occurring as a result of differences or changes in "settling rate", see FIG. 15, i.e., a dynamic condition; and not due to separation after the particles have arrived at their static bands, as shown in the classical separation bands of FIG. 14. b) Settling rate in a given medium is a function of particle size, shape, concentration, and density. It is also a function of fluid viscosity and is generally proportional to the term r ω², where r = distance of particles from spin axis; ω = spin angular velocity, which is selected to achieve dynamic settling rate separation of different particles. FIG. 15 is a plot of the population distribution for several formed elements of blood with respect to their settling rates divided by the multiplier r ω². Dilute concentrations are assumed and the Stokes approximation is used for these plots. There is good agreement with observation for all blood particles except concentrated red blood corpuscles (RBC) .

Note that in FIG. 15, the separation of platelets from other particles, as a function of settling rate, is much greater than in FIG. 14, which is a plot of the population distribution as in FIG. 15 but, in this case, with respect to particle density only, wherein the platelet band overlaps (i.e., is not spaced from) the lymphocyte particle band. The density bands of FIG. 14 illustrate the end point distribution of essentially fully settled particles. c) In accordance with the present invention, decanting occurs while particulate is settling out, and not after settling is completed into bands as shown in FIG. 14. d) Further, in accordance with the present invention, decanting can occur after some particles have settled out according to density to form distinct bands, however prior art centrifugal separators employ means to extract material from these formed "layers", which is radically different from and inferior to that which is used in the present invention partly due to the fact that in the present invention, decanting occurs in a direction parallel to the force field (i.e., in the direction of vector 300 in FIG. 11) , and not at an angle to the force field, thereby eliminating otherwise obligatory and destabilizing shear to achieve decanting. The latter effect reduces the effectiveness of all known prior art devices; and e) particles separate even while they are still physically mixed in suspension not otherwise possible when separating particles from "formed" layers.

As a consequence of these novel principles, it is possible to separate platelets from lymphocytes, RBC, and granulocytes in apparatus designed in accordance with the present invention by selecting the appropriate operating parameters as described above. Separation is in accordance with differences in settling rate, (FIG. 15) not density, (FIG. 14) . Decanting flow velocity and spin rate determine which components follow the several possible liquid fractions. These are fully selectable for many diverse applications as hereinafter described.

Application of these inventive principles to important blood component separation problems may be illustrated by first referring to FIGS. 13, and 14 which show a summary of published data on the population distribution of effective diameters and densities, respectively, of four important classes of human blood components, namely, red blood corpuscles (RBC), platelets, lymphocytes, and granulocytes. Except in regions of high concentration of components, the settling rate of that component in human plasma may be computed or at least characterized from these data using the Stokes drag equation:

V_s = D² ΔP r ω² 18 μ

V_s - settling velocity (cm/sec) ;

D - equivalent particle diameter (cm) ;

ΔQ - density difference, particle minus medium μ - viscosity, medium r (J² - g's of centrifugal force if g = g acceleration of gravity.

Equivalent particle diameter refers to that diameter that gives the experimentally observed or "correct" settling velocity according to the Stokes equation which is more useful to show how settling velocity varies with design and operating parameters r and ω, respectively. To this purpose calculations based on the Stokes equation are plotted in FIG. 15 as velocity per unit of centrifugal acceleration. Note that granulocytes, due to size as well as density, settle much faster than other particles, and the lot settles at least eight times faster than platelets.

Consequently, platelets which could not be separated from lymphocytes by density (see FIG. 14) , which is a very significant problem in achieving the desired lymphocytes-free platelets required for therapy, can be separated by settling rate using the teachings of the present invention. In the operation of the apparatus of FIGS. 3, 4, and 5, as a separator for blood components, liquid mixture M is blood which enters the upper portion of the rotating cylindrical assembly A through drive tube 104 and enters gap G. This form of the apparatus is particularly adapted for use as a collector of cell free or platelet rich plasma.

In the usual blood-bank setting a typical donor can be "bled" at a rate of about 100 ml/min. At an average hematocrit (Hct) of 45% (i.e., percent of whole blood by volume comprising cellular material) , 55 ml/min of the 100 ml/min is plasma, the suspending medium, which, incidentally, is over 90% water, the rest being dissolved complex organic molecules, lipids, and salts. Because the device illustrated in FIG. 3 is capable of decanting as much as 90% of the plasma, leaving the remainder to support fluidity of the concentrated portion, one can use, for purposes of this example, the removal, i.e., decanting, of 50 ml/min of plasma uniformly distributed over 50 cm² of porous flow distributing surface. Hence, the inward radial convective decanting velocity is 1.0 cm/min or

0.0167 cm/sec, that is, 50 ml/min divided by 50 cm².

For operating and design purposes, one might pick a settling rate for red blood corpuscles to be 0.02 cm/sec which exceeds the convective velocity by enough to prevent red blood corpuscles from approaching the porous surface. Referring to FIG. 15, one has

0.8 x 10^"7 x r ω² = 0.02 or r ω² = 2.5 x 10⁵. A practical example for the diameter of a blood film could be 1.5 inch. With an r ω² of 2.5 x 10⁵ the r.p.m. would be 3486 giving 255 g's of centrifugal force. Inserting these values into the Stokes equation and solving for V_s, the corresponding settling rate of the fastest platelets is only 0.2 x 10^"7 x 2.5 X 10⁵ = 0.005 cm/sec or far less than the convective velocity of 0.0167 by a factor of 0.30. Hence platelets are easily carried with the plasma to yield what is referred to as platelet rich plasma (PRP) but without, it should be noted, the presence of any leukocytes.

This calculation assumes a relatively low concentration of red blood corpuscles near the porous surface for the Stokes equation or, more to the point, a narrow well defined range of settling velocities to apply. The initial impregnable zone makes this possible as follows:

Suppose the blood film thickness is nominally 0.1 cm, which, as will be shown herein, is well within the allowable range. When the red blood corpuscles and lymphocytes travel about 0.05 cm, they are nearly fully settled out as together they comprise nearly half the volume of the initially whole blood. They travel nearly this entire distance, or about 0.04 cm in 2 seconds, in the absence of a radial convective velocity such as obtains over the initial impregnable zone. Since the feed rate of 100 ml/min is 1.667 ml/sec, a hold-up volume of 3-1/3 ml of blood film over this zone (which may, for example, be obtained by using a gap G thickness of 0.1 cm and a first encountered impregnable surface area of 33-1/3 cm²) will cause the blood to have a residence time of 2 seconds over the impregnable area and 6 seconds over the 50 cm² of porous surface 21. The latter value is due to the fact that half of the 100 ml/min is withdrawn resulting in an effective axial flow of 50 ml/min through a hold-up volume of 5.0 ml over 50 cm² of porous surface. The total residence time is 8 seconds in the operative blood film.

As just seen, the red blood corpuscles are nearly settled out before they even get to the porous surface 21 and settle even further thereafter. It is therefore reasonable to conclude that throughout the proximity of the porous surface, the concentration of red blood corpuscles is low and the Stokes settling rate is obeyed.

On the other hand, in 2 seconds, the platelets move away from the surface by, at most, 0.005 cm or barely 1/lOth of the distance they would otherwise travel to their settled position absent convective decanting. They retrace this entire displacement in less than l/3rd second after encountering radial inward convective flow upon first reaching the porous surface 21 and continue to be drawn through the porous flow distributing surface for the remaining 5.67 seconds of residence time over that surface.

These simple calculations demonstrate a device that yields nearly all the available plasma completely rich in platelets and absent all other cellular material. The size of the device for the application example indicates that it could easily have been made smaller say 1.0 inch diameter, and ω increased to 4270 r.p.m. to achieve the same effect. On the other hand, there is another use for the device wherein the larger device can be used to advantage, that iε, where one wants to either concentrate the collected platelet rich plasma in a second stage (see FIG. 12) or simply collect cell-free plasma.

In the latter case, assume again that the decantation rate is 50 ml/min over 50 cm² of porous surface or a convective velocity of 0.0167 cm/sec. For completely cell-free plasma the settling rate of platelets must exceed this value, say 0.02 cm/sec or greater. Then, because platelets are so εparsely concentrated to begin with, (i.e., their concentration has no affect on the validity of the Stokes calculation) one may use: r ω² x 0.5 x 10^"8 > 0.02 or, for a 2.0 inch diameter blood film, the r.p.m. must be 12,077. This rotation speed can be achieved with proper bearing design and balancing of the apparatus. By way of example, FIG. 16 illustrates the manner in which two separator apparatus A-l and A-2 of the type shown in FIG. 3 may be operated simultaneously to produce platelet concentrate (PC) using positive displacement peristaltic roller pumps commonly available for the pumping of blood or other sterile fluids.

In the operation of the apparatus of FIGS. 6, 6A, and 7 as a separator for blood components, liquid M would be blood which enters the upper portion of the rotating cylindrical assembly A through drive tube 104' and enters gap G'. This form of the apparatus iε particularly adapted for use as a rapid PC collector in certain surgical settings where high platelet loss is expected. If a fully therapeutic dose of platelets could be obtained from the surgical patient in the O.R. during surgical preparation and returned to that patient following surgery, significant clinical benefits (related to control of bleeding, i.e. clotting mechanisms) would be expected. H a v i n g reference to FIGS. 6, 6A, and 7, such a procedure can be carried out in one step in the apparatus shown. Because access to venous circulation is subεtantial in thiε caεe it iε poεεible to obtain, for example, about 300 ml/min of blood flow, at leaεt half of which, 150 ml/min, will be plaεma available for collection or processing. Of that, 30 ml/min will carry the concentrated platelets and 120 ml/min would be the cell-free plasma to be decanted first. Referring to FIG. 6, with blood entering gap G' at the top, such blood first encounters impervious entrance member 66' and then porous sectionε 21' and 170 in succession. For purposes of the present example, the area of porous surface 34' of porous element 21' is fixed at 200 cm². 120 ml/min of cell free plasma flowing radially inward through the porous element produces a radial convective velocity of 0.01 cm/sec, i.e. 2 ml/sec divided by 200 cm². In order to collect only cell-free plasma as the second liquid fraction M2, the Stokes velocity for plateletε must exceed the convective velocity by at leaεt 25%. In the preεent example r ω²x 0.5 x 10^"8 > 0.0125 represents the required condition. See FIG. 15. For a 2.0 inch diameter blood film the r.p.m. is 9548. Larger diameters or more porouε surface area would reduce the r.p.m. requirement.

The plasma remaining in zone E after decantation of cell free plasma from zone RC contains PC which can, in sequential decantation acrosε the outer surface of porous element 21' be largely removed from zσne E through porous ring 170 as third liquid fraction M₃, the remainder rejoining the concentrated material from zone 83' at pasεage 144' where the recombined material, now first fraction M.., is led away at conduit 86'. In order to prevent RBC or lymphocytes from being decanted along with the platelets it iε necessary to chooεe a convective radial velocity which exceedε the εettling velocitieε of platelets but is lesε than thoεe for RBC. Referring to FIG. 15 it iε seen that such a velocity would be adequately represented by r ω² x 2 x 10^'8 which, for the specified blood film and rotation rate, is 0.05 cm/sec. Given the decantation rate for M₃, (i.e. PC), that is, 30 ml/min or 0.5 ml/sec, the area of the exterior surface of porous ring 170 must be 10 cm² to enable the convective inward velocity of 0.05 cm/sec at that surface. Thus, the εecond porouε element 170 is much smaller than the first porous element 21 in area and axial extension. Referring again to Figs 6, 6A, and 7, yet another application of the invention has great utility in the washing and concentrating of red blood corpuscleε either in connection with frozen blood storage or, perhaps more importantly, for application to blood cell salvage in what is termed auto-transfuεion of the surgical patient's own blood during surgery. In the latter instance, it is the removal of excess heparin and return of concentrated protein and clotting factors for the purpoεe of clotting function regulation that haε the greateεt clinical importance. The proceεεing of high volumetric rateε can be uεeful. Cell waεhing should involve a concentration step followed by the simultaneous addition, separation and removal of wash solution, usually saline. If concentration precedes washing, the amount of wash solution required to displace cell stroma dispersed in the original suspending medium is minimized.

With particular reference to FIG. 6A, when cylindrical valve ring 182 is raised, a saline wash solution, indicated by flow arrows 189 from space 93' flows radially inwardly through porous rings 184 and 170, respectively, at a rate of flow effective to enable an inward radial velocity of controllable and uniform value acrosε the gap and into paεsages 174, space 176, and through discharge conduit 178 (see FIG 6) . The inward radial convective velocity associated with wash solution flow should be higher than that of prior flow acrosε surface 34' of porous element 21' in order to take full advantage of a sequential concentrating followed by a washing step intended to remove contaminants not otherwiεe permitted to follow with flow acroεs such surface. Using the criteria above deεcribed, assume the apparatus of the present invention is capable of removing 200 ml/min from a blood flow that would be typically 300 to 400 ml/min. Also, assume surface 34' has an area of 200 cm² so that the inward radial convective velocity iε 0.0166 cm/sec. In order not only to salvage most of the red blood corpuscles, especially the youngest and healthiest cells, which are generally the least dense, but further, to prevent smaller contaminants such as cell stroma which have settling velocitieε comparable to plateletε, from following with the protein comprising the second liquid fraction, one would again consult FIG. 15 where the minimum Stokes velocity for platelets is about 5 x 10^"9 x r ω² and εet this value at or above 0.02. Using a 2.0 inch diameter blood film, the required r.p.m. is 12,080. It remains only to use a higher radial velocity than that used to decant liquid from zone RC to decant wash solution, after deciding how much waεh εolution iε required, bearing in mind it can always be adjusted to fit whatever operating conditions prevail and that the example here is for maximal conditions. The operative parameter is about 10 ml/min of wash εolution for each 1.0 cm² of exterior and interior porous surface of porous elements 170 and 184, respectively, in order to maximize the washing effect (i.e. displacement of cell stroma) while still retaining all RBC in the first fraction. Because the pre-concentration step removed so much protein containing liquid which does not now have to be displaced by an approximately equal amount of wash solution, it is estimated that 20 to 50 ml/min of wash solution would be sufficient to proceεs about 100 ml/min of packed RBC. Using the upper value, the surfaces of porous elements 170 and 184 and could be as little aε 5 cm². However, if one doubleε the area and rate of waεh solution flow, the design iε virtually identical to that for the PC collector. It will be understood that operation of cylindrical valve ring 182 to block and unblock accesε of saline solution to porous elements 184, as shown in FIG. 6A, converts the FIG. 6 apparatus from PC collector to a cell washer and concentrator.

It is noted that the washing process operative in the example herein described displaceε rather than mixeε with the original suspending medium as in prior art cell washing methodε. Conεequently, washing is more thorough per unit volume of wash solution because, as in the compariεon between a εhower and a bath, the waεh solution is not substantially diluted. It is further noted that the crystalloid portion of the first separated protein containing liquid (i.e. plasma) can be removed as by ultra filtration in separate apparatus so that protein concentrate can also be returned to the patient. See FIG. 17.

Figure 17 is a schematic illustration of the operation of separator apparatus embodying the present invention which is convertible from a platelet concentrator to a cell washer/concentrator in accordance with principles of the invention. Whole blood or a solution containing blood components is delivered by pump P6. As a PC collector or cell washer/concentrator, pump P7 removes concentrated blood while pump P8 removeε cell-free and stro a-free liquid. The difference, that is, the volumetric pumping rate of P6 lesε the εum of rateε of P7 and P8 is a net flow which is PC or contaminated wash solution in the cases of platelet collector and cell washing/concentrating, respectively. In the former, cell-free plasma is recombined with concentrated blood to form reconstituted blood for return to the patient. In the alternate, cell-free liquid, containing plasma proteins and crystalloid material including water, follows a different path as indicated by the dotted lines at the left portion of FIG. 17 to a separate device incorporating membrane filtration to remove a large part of the water and crystalloid to form concentrated plaεma protein. The latter iε recombined with concentrated blood for return to the patient.

Saline circulation iε handled in each caεe aε follows: Pump P9 delivers saline to each end of the apparatus while pump P10 removes saline from the central port 190 simultaneously, and at a slightly slower volumetric rate, in the case of platelet collection, in order to force some saline acrosε the seals.

For cell washing, pump P10 is slowed further. Pump P10 removes saline from the apparatus A at a volumetric rate which is lesε than the volumetric rate of pump P9. The difference iε the volumetric rate at which εaline waεh εolution iε delivered acroεε the gap. It iε understood that fluids other than saline may be used for the purpose described above. When the cylindrical valve ring 182 (FIG.6) is raised, εaline solution enters the porouε ring 170 and outlet conduit 178 to waεh contaminants picked up by the saline solution for dispoεal, aε indicated by the dotted lines at the right portion of FIG. 17. Referring now to FIGS. 8 and 9, there is εhown the major components of a third form of apparatus embodying the present invention. This form of the invention is similar to the form of the invention εhown in FIGS. 3, 4, and 5, with the exception that the outer cylinder O iε provided with a porouε element 193 over a portion of itε axial extenεion. Like partε bear double primed reference numerals. The upper part of the apparatuε of the third form of the invention is provided with an electro-magnetic armature drive, like that shown in FIG. 3 arranged to effect rotation of the inner and outer cylinders I'' and O'' within stationary housing H''. A sterile solution is introduced into the conduit 91'' at top of drive support member 104'', and is discharged through an outlet conduit 92'' at the lower right-hand portion of housing H''. The liquid component M to be fractionated enters through the upper end of drive tube 104' ' and flows downwardly through the lower end of the drive tube into the entrances of radially outwardly extending pasεageε 196 formed between the upper walls of the inner and outer cylinder members. The radially outer ends of theεe paεsages empty into the upper end of gap G''. The lower end of the gap G^,/ iε in communication with the radially outer endε of radially extending paεsages 209 found between the lower walls of inner and outer cylinders I'' and O''. The radially inner ends of these pasεages are connected to the upper end of lower tube 150'' to discharge liquid fraction Ml through the lower end of the housing H''. A circular baffle 210 is arranged coaxially between support tube 41'' and inner cylinder I'' to define inner region R' ' . Inner region R" receives liquid fraction M2 which permeateε through porouε element 21'' of the inner cylinder I'' by a preεεure differential created between the gap G ' ' and inner region R' ' , as fully described hereinbefore. Passages 211 formed in the lower portion of the cylindrical aεεembly A' ' transfer liquid component M2 into a space 212 from which the liquid is discharged through the lower portion of housing H''. The lower portion of the houεing H'' is provided with bearings and O-ring bearing seals (not shown) through which a sterile, cooling solution is circulated as with the apparatus of FIGS. 3 and 6.

Assuming the liquid mixture is blood, for cell waεhing purpoεeε, blood entering drive tube 104'' at the top of the apparatuε uεually conεiεtε of concentrated cellε (i.e., 90+% cellε in a liquid medium or "carrier") . The blood cell concentrate iε led via flow path 207 to a gap entrance region 208 at the upper end of the separation gap G''. The gap is defined by interior surfaces of the inner and outer shell's axial sections. Sterile εolution, εuch aε a saline wash solution, enterε the upper and lower endε of houεing H'' by means of positive displacement volumetric pumping, follows a flow path through conduits designed to cool the bearings and seals and prevent blood component leakage and enters exterior region space 93'' between housing H'' and outer cylinder 0'' at a substantially uniform pressure throughout the exterior region. It further enters and crosseε (i.e., permeates) porous element 193 of the outer cylinder under the influence of the uniform pressure at a uniform volumetric rate over the entire exterior surface of such porous element, thus establishing an inward radial flow velocity of wash solution acrosε the axial extent of gap G" encloεed by porouε element 193, as described hereinbefore. The flux velocity is determined by the volumetric rate at which wash εolution croεεes porous element 193 divided by the inner εurface area 213 of εuch porouε element. The portion of waεh solution that enters space 93'' in excesε of that which crosses porous element 193 leaves the apparatus via port 92''.

Wash solution which crosses porous element 193 nearest the top of the gap enters the concentrated blood cell mixture at gap region 218 and operates to dilute the mixture (provided wash solution is miscible with carrier fluid) even aε the cells migrate radially outwardly under the influence of centrifugal force. The cells form a concentrated cell masε exterior to imaginary boundary 48''. Wash solution accumulates in gap region 219 and flows axially downwardly along with the concentrated cell mass in gap region 218. Gap region 219 is formed by impervious entrance element 216 which compriεes an axially extending portion of the inner shell at the latter's top, which portion iε tapered radially inwardly or away from boundary 48'' in the downward direction. Consequently, gap region 219 increaseε in thickneεε downwardly to accommodate the addition of wash solution to the mixture.

Wash solution enters gap region 220 at a flux velocity identical to that at which it enters region 218. However, wash εolution is simultaneously crosεing the outer surface of porous element 21'' at the same rate thus removing wash solution from gap region 221 as fast as it enters. The net effect is that wash εolution croεεeε the gap in a radially inward direction with no net accumulation of waεh εolution in the gap for the axial extent of the gap which iε defined and enclosed by the common axial overlap (i.e., axially co-extenεive portions) of porous elements 21" and 193 acting together. The axial extensionε of porouε elements 21'' and 193, where they do not face each other, serve purposes deεcribed separately herein. The gap-defining surfaces of the porous elements, where they face each other, remain concentric and parallel to boundary 48'' and to each other. The latter is the principal washing region of the gap.

The washing method is practiced by causing the wash solution flux velocity to be greater and less than the settling velocities of particles and/or cellε to be washed out and cells to be retained in the gap, respectively. Clearly, continuous phase will be washed out and virtually replaced with wash solution, provided the two liquids are miscible. If they are not, it is not poεεible to predict with certainty the behavior just described. Waεh εolution that was added to gap regions 218 and 219 is an excess volume not removed from gap region 221 and must flow axially toward exit region 209. However, the bottom portion of the outer cylinder is comprised of an impervious element 222 having an inner gap defining surface 223 for approximately the same axial extent aε the imperviouε upper element 216 of the inner shell. Impervious surface 223 is a continuation of porous element 193 wherefore wash solution flows out of gap region 224 without simultaneous replenishment from gap region 225. Consequently, porous surface 226 of element 21'' is tapered outwardly and downwardly toward boundary 48'' to account for the diminishing amount of continuous phase. Cells in gap region 225 become more concentrated until they reach the level of concentration at which they entered, thus canceling the dilution introduced in gap regions 218 and 219. Surface 226 is tapered outwardly and downwardly to boundary 48'' to aid in the flow of concentrated cells toward exit region 209.

The described method of dilution and re- concentration has the effect of increasing the distance between imaginary boundary 48'' and the inner shell's porous outer surface 226. The additional distance operates to maintain dilute concentration of cells in the mixture most proximate the porous "εeparation" εurface 226. The main concentration of cellε is radially exterior of boundary 48''.

The "environment" in which biologically valuable cells exist in gap regions 220 and 221 is unique in the present invention and, being suεtainable indefinitely, has not been so achieved in apparatus heretofore. The opposing forces on a biological cell, operative in the gap, are centrifugal force urging the cell to seek the inner porous surface of the outer shell and the drag force of radially inwardly flowing media tending to carry the cell toward the outer porous surface of the inner shell. All cells remaining in the gap find a point of balance between theεe forceε or "reεt" on the inner surface of the outer shell which contains pores through which the cells cannot pasε, not necessarily because the pores are too small in "εize" for cells (some of which are sufficiently deformable to enter pores smaller than their normal crosε-εection) to enter but, rather, becauεe the pore-open to pore-matrix solid volume is sufficiently low to subεtantially increaεe the waεh solution inward radial flux velocity obtaining within the porous element volume to overcome the outward radial migration velocity of all relevant cells. Again, "filtering", understood as a sieving process, is not the operative feature here, nor does the outer porous element operate as a physical barrier, absent the radially inward flow of wash solution, to effect separation of components.

For every collection of cells there exists a wash solution flow rate corresponding to an inward radial flux velocity in the gap in combination with a centrifugal force, determined in the apparatus by rp , that can easily be found experimentally to produce a cell concentration gradient most dilute near the inner shell and most concentrated near the outer shell wherein every cell is surrounded by wash εolution flowing continuously over its surface and wherein every cell no longer moves radially in either direction. In effect, the cells can be levitated and maintained in suspension even while the suεpending phaεe is being continuously ex changed. The cell mass exists under conditions which are analogous to a "liquified bed". Such apparatuε can εuperimpoεe a controlled axial flow or operate without axial flow by deleting gap regionε 218, 219, 224, and 225 along with the associated impervious elements. Once εtasis iε achieved, it can be maintained indefinitely. Waεh solution can carry nutrients and oxygen to the cellε and εimultaneouεly remove C0₂ and other waεte productε. It can be uεed to control temperature, enzymes and drug delivery to the cell. It is clearly a novel cell culturing method that does not subject the cell to the damaging effects of mechanical εhear and filtration or εubεtrate interaction as used in some prior cell culturing methods.

Furthermore, given the indefinitely extended time scale, separation as between cells with very closely matched sedimentation properties could be carried on very gradually over long periods of time. These novel techniques suggest promising new possibilities for further biological and medical research involving in vitro cell studieε. As a more immediate practical application of the cell washing method of the present invention, FIG. 19 illustrateε a simple pumping arrangement for washing stored blood using the εeparator apparatuε of FIGS. 8 and 9. Blood is pumped via pump P15 while washed cell concentrate is pumped via pump PI6 into a separate washed-blood bag. Becauεe of the normal time delay required for some chemicals εuch as ethylene glycol (typically used to store frozen blood) to leave the red blood corpuscleε (RBC) one would allow the firεt waεhed blood to re-enter the original εtorage bag, after all blood iε cycled through the cell waεher, by opening a valve between the two bags. After a period of time, the cycle is repeated for as many cycles as may be required to rid the εtored blood of objectionable chemical. The laεt cycle leaves the clean blood in the washed-blood bag. Pumpε P18 and P19 provide circulation and poεitive preεsure on the seals of the apparatus A5 for controlled leakage as described hereinbefore, but the difference in their volumetric rates, i.e., the presεure of pump P18 minuε the preεsure of pump P19 is now also uεed to provide the poεitive definite flow of waεh εolution. The volumetric rates of pumps P15 and P16 match at all times so that there is never any difference in blood volume. Consequently, all exceεs flow, i.e., flow from pump P18 minus flow from pump P19 must leave the separator apparatus A5 via the waste wash path.

Another use of the cell washing method and apparatus of FIGS. 8 and 9 permits a major departure in extracorporeal blood oxygenation not posεible heretofore. FIG. 20 illustrates such general method. The first stage employs a plaεma εeparator apparatuε A6 of the type shown in FIGS. 8 and 9 yields two outlet streams, one being highly concentrated RBC bearing de-oxyhemoglobin, that is, red cells poor in oxygen and εaturated with carbon dioxide, RBC-D, the other being moεt (90% to 95%) of the plaεma which iε returned to the patient. The concentrated RBC-D (not pumped in order to minimize hemolyεiε) , is led to a mixing chamber 400 through conduit 402 where it is thoroughly mixed with concentrated oxygen rich artificial blood particles, AB-0. The resulting mixture contains RBC bearing oxyhemoglobin, RBC-0, and spent AB saturated with C0₂ or AB-D. The mixing chamber output rate of flow iε determined by the rate of pump P20 (PR1) minus the rate of pump P21 (PR2) plus the rate of pump P25 (PR6) , which flow enters the cell washer A7 through conduit 403. The difference of pump rates, PR1 minus PR2 is determined by the hematocrit of the patient blood (i.e., cell volume fraction, Hct.). The rate of pump P25 (PR6) will necesεarily be related to that difference which correlates with the mass flow of hemoglobin, Hb., that is:

PR6 = f (PR1 - PR2) where f is some factor depending upon the AB product used.

The wash εolution flow rate iε the rate of pump P23 (PR4) minuε the rate of pump P24 (PR5) even though the wash solution is led serially through the plasma separator A6 (which draws off very little wash solution) before entering the cell washer A7, as shown in FIG. 20. Again, the concentrated RBC-0 is not pumped in order to protect the cells but pump P22 draws waste wash solution out of the syεtem. So long aε PR3 = PR6 + (PR4 - PR5) or PR3 = (PR4 - PR5) + f (PR1 - PR2) , the cellular output of RBC-0 from the cell washer A7 will match that of the plasma separator A6, i.e., PR1 - PR2. This preεerveε the RBC cell mass. The waεh rate PR4 - PR5 iε controlled to provide the neceεεary waεh-out of AB particles and suspending medium. In order to prevent repeated dilution of plasma protein with each pass of blood through the oxygenator, the waεh εolution in this example should be the patients own plasma or donated plasma. This is both posεible and practical because very little wash solution is required per pasε to remove AB particleε from the concentrated RBC cell mass in accordance with the method of the present invention. Finally, the RBC-0 is recombined with the plasma from the plasma separator to reconstitute the blood, i.e., plasma with, now, oxygen-rich, C0₂- regulated, RBC. The reconstituted whole blood must flow into sufficient positive presεure εo that neither pumps P21 nor P22 apply suction to the fluidε. Thiε provision is important to prevent fluid outgassing or cavitation. The suction side of pump P22 is the loweεt pressure in the syεtem, but it need not be below atmoεpheric if there is adequate reεiεtance in the patient return line 404. Having deεcribed the εyεte , it iε important to note that there are many varietieε of artificial blood generally conεiεting of particles of extraordinary solubility relative to gases εuch as 02 and C0₂. Many are very compatible with blood and organ tiεεueε. A typical example is Perfluorooctyl Bromide, PFOB, sold by Nippon Mektron Ltd. The ability of theεe particles to rapidly transfer oxygen and remove C0₂ from the body tiεεueε and blood cellε, with no apparent toxicity, iε well eεtabliεhed. Consequently, direct mixing of PFOB with concentrated RBC-D converts the latter immediately to RBC-0. The problem has always been separating the spent PFOB from the blood before returning the latter to the patient. Wholesale injection of PFOB into the patient is not approved.

Complete separation of entrained PFOB particles by filtration has not been possible because such particles are only about 0.1 micron in diameter. Separating them by prior centrifugation techniques is equally difficult because the particles have a specific gravity approaching 2.0 and would form a thin dense layer under the RBC which would defeat any reasonable method of skimming. On the other hand, the settling velocities of the particles are a εmall fraction of the εettling velocity of RBC due to their small size. The settling velocity of a PFOB particle is about 0.006 the settling velocity of an RBC. Consequently, they are easily washed out using the present invention.

In εummary, the oxygenator εystem of the preεent invention aε deεcribed immediately hereinabove includeε a plaεma εeparating εtep (εo as not to require separating PFOB particles from plasma) followed by a step which mixes RBC with AB, followed by a step which washeε AB and any other εtroma out of the RBC and finally recombination of clean, oxygenated RBC with the plasma for return to the patient. These steps run continuously and simultaneously in a closed extracorporeal circuit aε will be clear from FIG. 20. With reference to porouε elements disposed in the inner shells I, I', and I" of FIGS, l through 9, their necessary properties may be summarized aε follows. They should have:

1) pore diameters of sufficient size to pasε formed or particulate supernatant constituentε without reεtriction, but not substantially larger than that size; and

2) resiεtance to flow of the εupernatant fluid medium εo that preεεure loss experienced by supernatant crosεing the porous barrier is substantially greater than presεure loεε experienced by the remaining εuεpension as the latter flows from entrance to exit of the device; and

3) because axial movement of the blood within the blood film must be under conditions of viscouε boundary flow in order to remain εtable, the outermost layer of the porous cylinder or outer diameter εurface must be "wettable" by the suspending medium, hydrophilic in the case of blood. Examples of how these properties can be achieved in a practical device are as follows: First, it should be noted that the porous flow distributors are esεentially tubular and can be comprised of multi-layers of concentric cylinders of various properties which, acting in combination, achieve the desired results. Once the outer surface iε "wetted", the interior remainder of the porouε structure need not obey this condition (i.e., hydrophobia) which is one way of significantly increasing flow resiεtance.

It iε further deεirable, although not aε fundamentally required, that the outer diameter of the porouε cylinder have a high "open" to "solid" volume ratio or stated in other terms, a high concentration of pores per unit surface area. This condition should exist for a depth of at leaεt several pore diameters in order to prevent local concentration of εupernatant flow velocity at a microεcopic level. Once the supernatant becomes trapped within the porous structure and removed from the bulk blood film, the high open to solid ratio need not be maintained. Consequently, inner layers of the continuously open porous structure can be comprised of lower open to solid ratios or lower concentration of pores. This further adds to flow resiεtance without affecting the requiεite blood contact εurface properties of the porous flow distributor. Of course, simply adding layers, or pore structure thickness, further adds to flow resistance.

In εummary, three wayε to eεtabliεh the deεired flow reεiεtance of the porouε flow diεtributor are:

1. Decreasing the open to solid volume toward the interior of the pore structure. 2. Increasing the depth or thicknesε of the pore structure. 3. Changing the "philic" nature of the surface in contact with the suspenεion to a "phobic" condition toward the interior. In practice, one may achieve theεe alterations of pore structure properties by employing concentric multi-layers.

It should be understood with respect to the aforedescribed apparatus and method of the present invention, that the fluid film, (e.g., blood) being separated must be thin enough that secondary flows which would disturb settling may not arise. The only flows permitted are the bulk rotation of the fluid film trapped between two surfaceε rotating at the same angular velocity, the axial laminar streamline viscouε boundary layer flow within that film from top entrance to bottom exit, and the forced convective radial inward flow through one or more porous flow distributing surfaces.

In general, flow in a rotating syεtem iε characterized by very εtrong secondary flows that arise as a conεequence of the rotational effects. Briefly, the Coriolis force balanceε the centrifugal pressure force in any regionε of the fluid flow that are unaffected by viεcoεity εuch aε would obtain sufficiently distant from a wall or surface containing the flow. In these regionε where the flow iε termed "geoεtrophic", the fluid velocitieε are nearly zero becauεe the force reεultε in no driving forceε to move the fluid. Thus, if there is a global flow occurring in the system, the flow must occur in a region where viεcoεity can upεet the geostrophic force balance to allow a fluid velocity. In many syεtemε, viscoεity is important only near the walls of the rotating container and, as a result, the bulk of the flow occurs only in thin layers near the wall. The layer where viscosity is important iε called the Ekman layer when the wall εurface normal vector is approximately parallel to the axiε of rotation, and the Stewartson layer when the wall surface normal vector is nearly perpendicular to the axis of rotation. A non-dimensional parameter that can be used to measure the importance of these boundary flows is the Ekman number given by e_k = υ/ωh² where υ is the kinematic viscosity of the fluid medium, ϋ is the rotational angular velocity, and L is a system characteristic length measured along the rotation axis. The Ekman number iε a ratio of the viεcouε force to the Coriolis force in a rotating system. In syεtemε rotating at high εpeed, e_k, is typically very small, suggesting that geostrophic flow dominates the character of the flow. Using the viscosity of blood plasma and the approximate dimensions of a device suitable for εeparating plateletε, the Ekman number iε tabulated below for εeveral rotation εpeeds.

TABLE I Speed (rpm) Ekman Number (e_k)

5,000 1.71 x 10^"7

10,000 8.56 x 10^"8

15,000 5.70 X 10^"8

20,000 4.28 X 10^"8 It is evident that geostrophic flow will dominate in a typical centrifugal separator used heretofore with blood because the Ekman number is so small. Consequently, all significant fluid flow will occur in a region very close to the walls of the procesεing volume which goeε to the heart of the reaεon for keeping the blood film "thin" in the inεtant invention. Incidentally, this problem plagueε all prior continuous centrifugal εeparators attempting to decant formed layerε during rotation in a bulk or batch separator. Estimates of the thickness of the boundary layers where the flow actually occurs can be made baεed upon the Ekman number and the length εcale. The thickneεε of the Ekman layer (on wallε perpendicular to a vertical axis of rotation, i.e., horizontal surfaces) is estimated by: t_ck = (ek)⁰-⁵ L The thickness of the more important Stewartson layer (on walls parallel or at some small angle to a vertical axis of rotation, i.e., vertical or nearly vertical walls) iε eεtimated by: t_st = (ek)⁰-²⁵ L Using these formulas, the estimated thicknesε of the layerε are indicated in Table II below:

TABLE II speed (rpm) t_£k (horizontal) t_st (vertical) 5,000 0.0019 inch 0.096 inch 10,000 0.0014 inch 0.081 inch

15,000 0.0011 inch 0.073 inch

20,000 0.0010 inch 0.068 inch

Clearly the blood film perpendicular to the rotation axis, that is, blood flow along horizontal surfaces in a vertically oriented device, are subject to instabilitieε and geoεtrophic flow. More important, however, is the fact that the operative separation chamber, namely, the "thin" blood film flowing parallel to the vertically oriented rotation axis iε within the Stewartson boundary layer thickness provided the blood film is limited to those thicknesεes listed in Table II under t_st. Theεe valueε define maximum claimed blood film thickness, at least for the entering whole blood.

Aε the blood becomes more concentrated and viscosity increases toward the blood exit, dramatically in some cases, one may wish to permit a thickening of the blood film by aε much as 20% to 30%, relative to the entrance value. It is noted that film thickness varieε only aε the 1/4th root of viεcoεity.

For moεt applicationε a blood film thickness of 0.060 inch at the whole blood entrance is adequate for the purpose of limiting axial preεεure gradients, even for flowε on the order of 1 to 2 literε/min.

Provided that the blood film thickness for the entire volume, where layering and separation is intended to occur, iε within the Stewartεon layer, aε defined herein, no geoεtrophic region with zero flow velocity will exiεt. Conεequently, undeεirable secondary flows are very unlikely. In short, the radial thicknesε of the proceεεing chamber iε εo εmall that the entire region can be thought of as a boundary flow. As a check on the potential for instabilities due to the forced convective flow radially inward toward the porous flow diεtributor, it is appropriate to examine the influence of inertial effects in rotating systems. This can be estimated by computing the Rossby number given by:

where υ iε a characteristic velocity in the plane perpendicular to the axis of rotation. Using the estimated convective velocity of 0.0167 cm/sec, the Rossby number was calculated for radial plaεma flow in whole blood as shown in Table III below:

TABLE III speed (rpm) R 0 5,000 6.4 X 10^"7

10,000 3.2 X 10^"7

15,000 2.1 X 10^"7

20,000 1.6 X 10^"7

Even at 20 times the convective velocity, as may obtain in surgical cell εalvage and concentration, and aε iε evident in Table III, the Rossby number is very small indicating that inertial effects are much smaller than Coriolis effects in this flow. Hence, the convective velocity does not contribute inεtabilitieε even approaching thoεe which limit the thickneεε of the Stewartεon layer. Other inεtabilitieε which might give riεe to inertial waveε could occur at the entrance to the proceεεing chamber but are quickly damped due to supporting vanes which separate walls defining entry to the blood film, and to the stabilizing influence of the initial impregnable zone which quickly eεtabliεhes the Stewartson viscouε boundary flow.

Finally, it iε important to conεider the influence of rotation on the εettling of the particleε in question. This will be done for platelets and red blood corpuscles.

The first isεue iε the particle Taylor number, T . Thiε parameter iε a meaεure of the ratio of the εize of the particle compared to the thickneεs of the Ekman layer, or, alternatively, the ratio between the Stokes drag force on the particle and the Coriolis force.

The particle Taylor number iε given by:

where a iε the particle radius. The particle Taylor numbers for platelets and red corpuεcles are listed below in Table IV.

TABLE IV εpeed (rpm) t (platelets) t (RBC)

5,000 1.09 X 10^'4 7.08 X 10^"4 10,000 2.18 X 10^"4 1.41 X 10^"3

15,000 3.27 X 10^"4 2.12 X 10^'3

20,000 4.36 X 10^"4 2.83 X 10^"3 If the particle Taylor number is large, which is not the case here, rotational effects can strongly influence the Stokes drag on the particle. This would have introduced error in the calculation of settling rates presented in FIG. 5(c), which can now be εafely relied upon. In the caεe of red blood corpuεcleε, the Stokeε drag iε εtrongly dependent upon particle concentration, εuch aε he atocrits typical of whole blood. However, it will be seen in specific examples to follow that red corpuscleε sediment very quickly leaving only platelet rich plasma for which particle concentration effects on the Stokes drag is minimal due to very low concentration of platelets on a volume percentage basiε. A εecond non-dimenεional parameter related to the settling of a particle is the time ratio λ. The time ratio is the ratio of separation time for a particle to the spin-up time for that particle. The spin-up time is the time for the particle to reach the syεtem rotational speed after it enters the rotating syεtem. The time ratio is given by: λ = e_k°-⁵/eT_c (L/r) e iε the denεity ratio defined by:

where ρ_c iε the denεity of the continuous phase, in this case, plasma, and ρ_d iε the density of the diεperεed phaεe, namely, plateletε or red blood corpuscles. Table V below lists the time ratios for these particles:

TABLE V speed (rpm) λ (platelets) λ (RBC) 5,000 13.0 1.20

10,000 4.6 0.43

15,000 2.5 0.22

20,000 1.6 0.15

This indicates that the separation of either red blood corpuscleε or plateletε is a relatively faεt proceεε. In fact, it may happen so quickly that the cells are not up to the rotational speed of the device (for λ << 1.0) . This would be a problem for red corpuscles at very high speedε. Potentially, red blood corpuεcleε could εediment to the outer wall before they have reached approximately the εame rotational εpeed as the wall. This could tend to cause high shear on the red blood corpuεcles with the posεibility of hemolyεiε which muεt be avoided. Therefore, conduitε and vanes at the blood entrance to the device serve not only to εpace and εupport the wallε defining the blood film but are indeed crucial in the deεign to force, especially the red blood corpuscles to spin-up to match device rotational speeds. Supporting vanes are both shaped and angled in such a manner as to accomplish the imparting of system rotational εpeed to the red blood corpuscles with a minimum of trauma, i.e., εhearing of the red blood corpuεcles among the spin-up εurfaceε. Aε εeen in Table V, the time ratio for plateletε is about 10 times that for red blood corpuscles and, consequently, red blood corpuεcleε εediment about 10 times faεter than plateletε as was εeen in the εeveral examples previously presented. In summary: 1) The appropriate blood film thickness has been computed and must be limited by viεcouε boundary flow or the Stewartεon layer.

2) Inertial inεtabilities are small, particularly in connection with the forced convection which decants the supernatant, i.e., the Rosεby number iε very εmall.

3) Turbulence iε not an issue because the Reynolds number iε leεε than 10 in this system.

4) The particle Taylor number indicates that, apart from concentration effects which influence only red blood corpuscles, the Stokes model for the drag on a particle iε valid.

5) The time ratio is εuch that the εettling of particles can be relatively fast in compariεon with their εpin-up time so that entry vanes are essential to force particle spin-up and εhould be deεigned with a minimum of turbulence and εhearing of the particles.

In practicing the present invention, the details may be replaced with other technically equivalent elements; furthermore, the materials used, and the shapes and dimensionε, may be any selected ones to meet individual requirements without departing from the scope of the present invention.

Claims

WE CLAIM :

1. Apparatus for fractionating a liquid mixture substantially into first and second liquid fractions whose components have settling velocities which are under identical conditions of centrifugal force, higher and lower, respectively, than a selected value, said apparatus comprising: a rotary asεe bly for rotation about a central axiε and including inner and outer shells defining therebetween an axially extending tubular gap and an inner region radially inwardly of said inner shell; said inner shell including a flow resistive porous structure to provide for permeation therethrough of said second fraction at a radially inward entrance velocity below εaid εelected value and to reεtrict flow of the reεidual of εaid fluid mixture therethrough; an inlet for introducing εaid liquid mixture into said gap at one axial end thereof; a frame for mounting said frame for rotation about said axis; a drive for rotating the assembly at a predetermined fractioning speed effective to impart centrifugal force to said liquid mixture within said εufficient gap to cauεe εaid εecond fraction to flow radially inwardly through said porous structure and to prevent said first liquid fraction from permeating said porous structure; and an outlet for drawing said firεt fraction from said inner region.

2. The apparatus of Claim 1 that includes: a residual outlet at one axial end of said gap for withdrawing said second fraction therefrom.

3. The apparatuε of Claim 1 that includeε: a pump for controlling preεsure in said gap to control the pressure differential acroεε said first porous structure.

4. The apparatus of Claim 3 wherein: said pump is operative to maintain the pressure of said liquid in said gap at a predetermined pressure and wherein said apparatus includes: a tank enclosing said asεembly and cooperating therewith to form a pressure chamber; encapsulating fluid in said preεεure chamber; and meanε for maintaining the pressure of said encapεulating fluid in εaid pressure chamber at a selected pressure higher than said predetermined presεure.

5. The apparatus of Claim 4 for use in fractioning a said mixture including blood components and wherein: said encapsulating fluid is sterile.

6. The apparatus of Claim 1, wherein: said inner shell includes an imperviouε cap between said inlet and said inner porous structure, the outside surface of said cap being tapered gradually radially inwardly and away from said inlet to gradually increaεe the annular flow area in εaid gap.

7. The apparatuε of Claim 1 for fractionating a liquid mixture of the type having a predetermined ratio of firεt to εecond fractions and which, during rotation of said asεembly at said predetermined fractioning speed, builds up an inner layer on the outer surface of said porous structure leaving an outer layer including said first fraction and wherein: said asεembly iε axially elongated; and said inner shell adjacent said mixture inlet tapers radially inwardly and away from εaid inlet at an angle to reduce εaid outer εurface thereof sufficiently so that as said drive rotates said aεεembly at said predetermined fractionating speed said inner and outer layers will flow axially at substantially the same axial velocity.

8. The apparatus of Claim 1 for fractionating a blood mixture and wherein: said inner and outer shells are concentric to form said gap with a radial thickness of between at least 5 and 150 millimeters.

9. The apparatus of Claim 1 for treating a blood mixture including plaεma and wherein: εaid porouε εtructure is constructed for, at said predetermined fractioning speed, permeation therethrough of said plasma.

10. The apparatus of Claim 1 for treating said second fraction with a selected fluid and wherein: said porous εtructure is conεtructed for, at said predetermined fractioning speed, permeation therethrough of said selected fluid; and said outer εhell includes an inlet for introduction of εaid selected fluid for flowing through said second fraction.

11. The apparatus of Claim 1 for fractionating a liquid mixture of blood with plasma defining εaid second fraction and red blood cells defining said firεt fraction and wherein: εaid porouε εtructure iε conεtructed for, at said predetermined fractioning speed, permeation therethrough of said plaεma.

12. The apparatuε of Claim 11 for drawing blood from a patient and which further includeε: an inlet conduit for connecting εaid inlet to a patient'ε blood circulatory εyεtem for drawing said blood therefrom.

13. The apparatus of Claim 11 wherein: said rotary assembly includes a residual outlet for withdrawing εaid red blood cellε aε εaid second fraction from said gap; and a blood outlet conduit connected to εaid reεidual outlet for withdrawing εaid red blood cellε therefrom for infuεion back into εaid patient'ε blood circulatory εyεtem.

14. The apparatuε of Claim 1 for fractionating a liquid mixture, including blood componentε εuεceptible to forming packed componentε which are reliquefiable by a selected fluid and that includeε: rotary εeal meanε configured to allow for εome fluid seepage of said selected fluid; an encapsulating houεing around said outer shell to be presεurized by εaid εelected fluid to a pressure sufficient to cause said selected fluid to seep through said εealε into εaid gap to reliquefy said packed components which might form in said assembly.

15. The apparatuε of Claim 1 wherein: εaid porous εtructure is axially elongated and εaid apparatuε includeε, and iε configured to, and has a thickness and pore diameter effective to, when said fluid mixture fully covers the εurface thereof and a predetermined pressure drop is formed radially across said porous structure, establish a predetermined rate of permeation; and a pump coupled with said fluid mixture inlet for maintaining said predetermined presεure drop across said porous structure.

16. The apparatus of Claim 1 that includes: a tank encloεing εaid assembly and cooperating therewith to form a pressure chamber; bearings mounting said assembly from said tank; cooling fluid in said presεure chamber; and circulation meanε for circulating εaid cooling fluid in heat exchange relationεhip with εaid bearings.

17. The apparatus of Claim 1, wherein: said rotary asεembly iε axially elongated, εaid inlet diεpoεed at one end of εaid gap and that includes: a residual outlet at the oppoεite end of εaid gap; and inlet and outlet pumps connected with the reεpective said inlet and residual outlet to control the rate of fluid flow through said gap.

18. The apparatus of claim 1 for fractionating said fluid mixture of the type having a predetermined ratio of εaid first to second fractions and predetermined viεcosity characteristicε and wherein: εaid gap is axially elongated, said inlet disposed at one end thereof and that includes a residual outlet disposed at the opposite end thereof;

said inner shell and porouε εtructure iε so configured that, at said predetermined fractionating speed, said asεembly is operative to form an outer layer including said first fraction in said gap surrounding an inner layer including said second fraction on the outer surface of said inner shell; and a pump connected with said inlet for controlling flow of said mixture at a predetermined steady flow rate axially through said gap to cause said inner and outer layers to flow axially at substantially the same velocity.

19. The apparatus of Claim 1 for treating said second fraction with a treatment solution and wherein: said assembly is axially elongated with said inlet diεpoεed at one end of εaid gap and that includeε: an outlet disposed at the opposite end of said gap; a porous treatment εtructure in said inner shell spaced axially toward said outlet from said inner porouε structure for flow of said first fraction axially in said gap past the outer surface thereof; and a treatment solution tank εurrounding εaid outer shell and formed with a treatment port for introducing said treatment solution into said gap adjacent said porous treatment structure whereby said treatment solution may be introduced to said tank to flow radially inwardly over said second fraction of said liquid mixture to pass radially inwardly through said porous treatment εtructure.

20. The apparatus of Claim 19 that includes: valve means for controlling flow of said treatment solution to said wash port.

21. Apparatus of Claim 1 for use with an encapsulating fluid compatible with said fluid mixture and that includeε: a pump for maintaining the preεεure in εaid gap at a predetermined pressure; sealε for εealing εaid aεεembly againεt uncontrolled eεcape of εaid fluid mixture and conεtructed to, upon application of εaid encapεulating fluid at a selected pressure, provide for seepage inwardly therethrough to flow said fluid mixture away from said sealε; an encapsulating tank surrounding εaid aεεembly for receipt of εaid encapsulating fluid; and a pump connected with said tank for applying said encapsulating fluid at said selected pressure thereto.

22. The apparatus of Claim 1 that includes: means for presεurizing the liquid mixture within εaid gap during εuch rotation of εaid rotary aεεembly to poεitively compel flow of εaid second fraction radially inwardly through said porouε structure into said inner region.

23. The apparatus of Claim 1 for fractionating a liquid mixture containing red blood cells and plasma and wherein: εaid porouε εtructure at εaid fractionating speed provides flow resistance to radial inward flow of said red blood cells while allowing said plasma to flow freely therethrough.

24. A method for fractionating a liquid mixture into first and εecond liquid fractionε and including the following steps: selecting an asεembly of the type including inner and outer εhellε conεtructed for equal angular velocity rotation about a common central axiε, said inner shell including a porous structure for at least a part of its extent and a space interior of said porous structure; introducing said liquid mixture into εaid gap between facing surfaces of said inner and outer shells; compelling flow of said second fraction through said porous structure; and rotating said shells at a speed effective to prevent the first liquid fraction from inclusion in said flow causing said first and second fractions to form on opposite εides of said porous structure.

25. The method of claim 24 that includes: distributing the flow of said mixture substantially uniformly over the outer surface of said porous structure.

26. The method of claim 25 that includes: diεtributing said flow over said porous structure to cause the radial inward velocity of entry of said flow into said porous structure to be substantially uniform over its outer surface.

27. A method for fractionating a liquid mixture substantially into firεt and εecond liquid fractionε whoεe components have settling velocities which are, under identical conditions of centrifugal force, higher and lower, respectively, than a selected value, which includes the stepε of: selecting an assembly of the type having inner and outer εhellε defining therebetween an axially extending tubular gap, and a radially inward inner region, the inner shell including an axially extending porous structure dispoεed radially outward of said inner region for permeation therethrough of said second fraction at a radially inward entrance velocity substantially equal to said selected value and to resist radial inward flow of the residual of said fluid mixture; introducing said liquid mixture at a predetermined rate into said gap at one end thereof; rotating said aεεembly at a predetermined fractioning speed effective to impart sufficient centrifugal force to said liquid mixture within said gap to apply a radially outward velocity to said first liquid fraction to resiεt permeation of εaid porouε εtructure; and preεεurizing εaid liquid mixture within εaid gap during εuch rotation of εaid aεεembly to compel flow of said second liquid fraction radially inwardly through said porous structure into said inner region while retaining said first fraction in said gap.

28. The method of Claim 27 that includes: rotating εaid aεεembly at a sufficient speed and preεsurizing said gap with sufficient presεure to maintain a portion of εaid fluid mixture, including said first fraction, flowing along the exterior surface of said inner shell in a subεtantially uniform flow pattern.

29. The method of Claim 28 that includes: rotating said assembly at a sufficient speed and maintaining said preεεure in εaid gap at a sufficient level to maintain said residual of said fluid mixture flowing along said inner shell in a sufficiently thin layer to maintain laminar axial flow throughout the depth thereof.

30. Fractionating apparatus for treating a selected fraction suspended in a suεpension medium of a liquid mixture with a selected treatment solution, said treatment solution and selected fraction formed by respective components having settling velocities which are, under identical conditions of centrifugal force, higher and lower, respectively, than a selected value, said apparatus comprising: an axially elongated rotary aεεembly adapted for rotation about a central axiε, εaid assembly having inner and outer shellε defining therebetween an axially extending tubular gap and an inner region diεpoεed radially inwardly of εaid inner shell; εaid inner εhell including a flow reεiεtant porous structure for at least a part of its axial extension to provide for permeation therethrough of said treatment solution and to preεent resiεtance to radially inward flow of εaid εuspension fluid; a liquid mixture inlet to one end of said gap; a treatment εolution inlet for introducing said treatment solution into the radially outward portion of said gap; a frame rotatably supporting said rotary assembly for rotation about said axis; a drive coupled with said rotary assembly to rotate said assembly at a predetermined fraction speed sufficient to, when said mixture is introduced to said gap through said mixture inlet and said treatment solution is introduced through said solution inlet, apply a settling velocity to said predetermined fraction sufficient to hold it suεpended within said gap while said treatment solution iε introduced through said treatment εolution inlet cauεing it to flow radially inwardly through εaid εuεpenεion fluid and second fraction.

31. The apparatus of Claim 30 for use in washing a mixture of the type including red blood cells with a saline wash solution and that includes: a saline tank encapsulating said rotary assembly and formed with said treatment solution inlet port; and a treatment pump to pump said saline treatment solution into said tank and through εaid treatment solution inlet port.

32. The apparatus of Claim 30 wherein: said inner shell is formed adjacent said fluid mixture inlet with a frusto-conical shape to slope gradually radially inwardly and away from said fluid mixture inlet to gradually increase the radial thicknesε of said gap.

33. The apparatus of Claim 30 that includes: bearings interposed between said housing and assembly; and circulation means connected with said treatment solution inlet for flowing said treatment solution in heat exchange relationεhip with said bearings.

34. Fractionating apparatuε for fractionating a firεt fraction suspended in a suspension medium of a liquid mixture including a second fraction, εaid fractions formed by respective components having settling velocities which are, under identical conditions of centrifugal force, higher and lower, respectively, than a selected range of settling velocities, εaid apparatuε comprising: an axially elongated rotary asεembly adapted for rotation about a central axiε, said assembly having inner and outer shellε defining therebetween an axially extending tubular gap and an inner region disposed radially inwardly of said inner shell; said inner shell including a flow reεistant porous εtructure for at leaεt a part of itε axial extension to provide for permeation therethrough of said first fraction and to present reεistance to radially inward flow of the remainder of said fluid mixture; a liquid mixture inlet to one end of said gap; a frame rotatably εupporting εaid rotary aεεembly for rotation about said axis; and a drive coupled with εaid rotary aεεembly to rotate said assembly at a predetermined fractioning speed sufficient to, when εaid mixture iε introduced to εaid gap through εaid mixture inlet, apply εettling velocitieε in εaid εelected range of εettling velocitieε to hold said second fraction in said gap while flowing said first fraction through said flow resistant porous εtructure into said inner region. AMENDED CLAIMS

[received by the International Bureau on 17 February 1994 (17.02.94); original claims 1,2,15,18,24,27-30,32 and 34 amended; other claims unchanged (13 pages)]

1. Apparatuε for fractionating a liquid mixture εubstantially into firεt and second liquid . fractions whose components have settling velocitieε which are under identical conditions of centrifugal force, higher and lower, respectively, than a selected value, said apparatus comprising: a rotary assembly for rotation about a central axis and including concentric inner and outer shells having a common axis defining therebetween an axially extending tubular gap and an inner region radially inwardly of said inner shell; said inner shell including a flow resistive porouε structure to provide for permeation therethrough of said second fraction at a radially inward entrance velocity substantially uniform over the outer surface of said inner porous structure and below said selected value and to restrict flow of the residual of said fluid mixture therethrough; an inlet for introducing said liquid mixture into said gap at one axial end thereof; a frame for mounting said frame for rotation about said axis; a drive for rotating the assembly at a predetermined fractioning speed effective to impart centrifugal force to said liquid mixture within said sufficient gap to cause said second fraction to flow radially inwardly through said porous structure and to prevent said first liquid fraction from permeating said porous structure; and an outlet for drawing said εecond fraction from said inner region.

2. The apparatus of Claim 1 that includes: a residual outlet at one axial end of εaid gap for withdrawing a residual liquid fraction. 3. The apparatus of Claim 1 that includeε: a pump for controlling pressure in said gap to control the pressure differential acrosε εaid firεt porouε structure.

4. The apparatus of Claim 3 wherein: said pump is operative to maintain the pressure of said liquid in said gap at a predetermined pressure and wherein said apparatus includes: a tank enclosing said assembly and cooperating therewith to form a pressure chamber; encapsulating fluid in said pressure chamber; and means for maintaining the pressure of said encapsulating fluid in said pressure chamber at a selected pressure higher than said predetermined pressure.

6. The apparatus of Claim 1, wherein: said inner shell includes an impervious cap between εaid inlet and εaid inner porous structure, the outside surface of said cap being tapered gradually radially inwardly and away from said inlet to gradually increase the annular flow area in said gap.

7. The apparatus of Claim l for fractionating a liquid mixture of the type having a predetermined ratio of first to second fractions and which, during rotation of said assembly at said predetermined fractioning speed, builds up an inner layer on the outer surface of said porous structure leaving an outer layer including said first fraction and wherein: said assembly is axially elongated; and said inner shell adjacent said mixture inlet tapers radially inwardly and away from said inlet at an angle to reduce εaid outer surface thereof sufficiently so that as said drive rotates εaid aεεembly at said predetermined fractionating speed said inner and outer layers will flow axially at substantially the same axial velocity.

9. The apparatus of Claim 1 for treating a blood mixture including plasma and wherein: said porous structure is constructed for, at said predetermined fractioning εpeed, permeation therethrough of said plasma.

10. The apparatus of Claim 1 for treating said second fraction with a selected fluid and wherein: said porous structure is constructed for, at said predetermined fractioning εpeed, permeation therethrough of said selected fluid; and said outer shell includes an inlet for introduction of εaid εelected fluid for flowing through said second fraction. 11. The apparatus of Claim 1 for fractionating a liquid mixture of blood with plasma defining said second fraction and red blood cells defining εaid first fraction and wherein: said porous structure is constructed for, at said predetermined fractioning speed, permeation therethrough of said plasma.

12. The apparatus of Claim 11 for drawing blood from a patient and which further includes: an inlet conduit for connecting said inlet to a patient's blood circulatory system for drawing said blood therefrom.

13. The apparatus of Claim 11 wherein: said rotary assembly includes a residual outlet for withdrawing said red blood cells as said second fraction from said gap; and a blood outlet conduit connected to said residual outlet for withdrawing said red blood cells therefrom for infusion back into said patient's blood circulatory system.

14. The apparatus of Claim 1 for fractionating a liquid mixture, including blood components susceptible to forming packed components which are reliquefiable by a selected fluid and that includes: rotary seal means configured to allow for some fluid seepage of said selected fluid; an encapsulating housing around said outer shell to be pressurized by said selected fluid to a pressure sufficient to cause said selected fluid to seep through said seals into said gap to reliquefy εaid packed componentε which might form in εaid assembly.

15. The apparatus of Claim 1 wherein: said porous structure is axially elongated and said apparatus includes, and is configured to, and haε a thickneεε and pore diameter effective to, when εaid liquid mixture fully covers the surface thereof and a predetermined pressure drop is formed radially across said porous structure, establish a predetermined rate of permeation; and a pump coupled with said fluid mixture inlet for maintaining said predetermined pressure drop across said porous structure.

16. The apparatus of Claim 1 that includes: a tank enclosing εaid assembly and cooperating therewith to form a presεure chamber; bearingε mounting said asεembly from said tank; cooling fluid in said pressure chamber; and circulation meanε for circulating εaid cooling fluid in heat exchange relationεhip with εaid bearingε.

17. The apparatus of Claim 1, wherein: said rotary assembly is axially elongated, said inlet disposed at one end of said gap and that includes: a residual outlet at the opposite end of said gap; and inlet and outlet pumps connected with the respective said inlet and residual outlet to control the rate of fluid flow through said gap. 18. The apparatus of claim 1 for fractionating said liquid mixture of the type having a predetermined ratio of said first to second fractions and predetermined viscoεity characteriεtics and wherein: said gap is axially elongated, said inlet disposed at one end thereof and that includes a residual outlet dispoεed at the opposite end thereof; εaid inner εhell and porous structure is so configured that, at said predetermined fractionating speed, said assembly is operative to form an outer layer including said first fraction in said gap surrounding an inner layer including said second fraction on the outer surface of said inner shell; and a pump connected with said inlet for controlling flow of said mixture at a predetermined steady flow rate axially through said gap to cause said inner and outer layers to flow axially at substantially the same velocity.

19. The apparatus of Claim 1 for treating said εecond fraction with a treatment solution and wherein: said assembly is axially elongated with said inlet disposed at one end of said gap and that includes: an outlet disposed at the opposite end of said gap; a porous treatment structure in said inner shell spaced axially toward said outlet from said inner porous structure for flow of said first fraction axially in said gap past the outer surface thereof; and a treatment solution tank surrounding said outer shell and formed with a treatment port for introducing said treatment solution into said gap adjacent said porous treatment εtructure whereby εaid treatment εolution may be introduced to said tank to flow radially inwardly over said second fraction of said liquid mixture to pass radially inwardly through said porouε treatment structure.

21. Apparatus of Claim 1 for use with an encapsulating fluid compatible with said fluid mixture and that includes: a pump for maintaining the pressure in said gap at a predetermined presεure; seals for sealing said assembly against uncontrolled escape of said fluid mixture and constructed to, upon application of said encapsulating fluid at a selected pressure, provide for seepage inwardly therethrough to flow said fluid mixture away from said sealε; an encapsulating tank surrounding said asεembly for receipt of said encapsulating fluid; and a pump connected with said tank for applying said encapsulating fluid at said εelected pressure thereto.

22. The apparatus of Claim 1 that includeε: meanε for preεεurizing the liquid mixture within εaid gap during εuch rotation of εaid rotary assembly to positively compel flow of said second fraction radially inwardly through said porous structure into said inner region. 23. The apparatus of Claim 1 for fractionating a liquid mixture containing red blood cells and plasma and wherein: said porous structure at said fractionating speed provides flow resistance to radial inward flow of said red blood cells while allowing said plasma to flow freely therethrough.

24. A method for fractionating a liquid mixture into first and second liquid fractions and including the following steps: selecting an assembly of the type including inner and outer shells constructed for equal angular velocity rotation about a common central axis, said inner shell including a porous structure for at least a part of its axial extent to resist radial inward flow of said liquid mixture substantially uniformly over the outer surface of said porous structure; introducing said liquid mixture into said gap between facing surfaces of said inner and outer shells; compelling flow of said second fraction through said porous structure; and rotating said shells at a speed effective to prevent the first liquid fraction from inclusion in said flow causing said first and second fractions to form on opposite sides of said porous structure.

25. The method of claim 24 that includes: distributing the flow of said mixture εubεtantially uniformly over the outer surface of said porous structure. 26. The method of claim 25 that includes: distributing said flow over said porous structure to cause the radial inward velocity of entry of said flow into said porous εtructure to be substantially uniform over its outer surface.

27. A method for fractionating a liquid mixture εubstantially into first and second liquid fractionε whoεe componentε have εettling velocities which are, under identical conditions of centrifugal force, higher and lower, respectively, than a selected value, which includes the εtepε of: selecting an assembly of the type having inner and outer shellε defining therebetween an axially extending tubular gap, and a radially inward inner region, the inner shell including an axially extending porous εtructure disposed radially outward of said inner region for permeation therethrough of said second fraction at a radially inward entrance velocity substantially uniform along said axial extension equal to said selected value; introducing said liquid mixture at a predetermined rate into said gap at one end thereof; rotating said assembly at a predetermined fractioning speed effective to impart sufficient centrifugal force to said liquid mixture within said gap to apply a radially outward velocity to said first liquid fraction to resist permeation through said porous structure; and pressurizing said liquid mixture within said gap during such rotation of said asεembly to compel flow of εaid εecond liquid fraction radially inwardly through εaid porouε εtructure into said inner region while retaining said first fraction in said gap. 28. The method of Claim 27 that includes: rotating said assembly at a sufficient speed and pressurizing said gap with sufficient pressure to maintain a portion of said liquid mixture, including εaid first fraction, flowing along the exterior surface of said inner shell in a substantially uniform flow pattern.

29. The method of Claim 28 that includes: rotating said assembly at a sufficient speed and maintaining said pressure in said gap at a sufficient level to maintain said residual of said liquid mixture flowing along said inner shell in a sufficiently thin layer to maintain laminar axial flow throughout the depth thereof.

30. Fractionating apparatus for treating a selected fraction suspended in a suspension medium of a liquid mixture with a residual and a selected treatment solution, said treatment solution and selected fraction formed by respective components having settling velocities which are, under identical conditions of centrifugal force, higher and lower, respectively, than a selected value, said apparatus comprising: an axially elongated rotary assembly adapted for rotation about a central axis, said assembly having inner and outer shells defining therebetween an axially extending tubular gap and an inner region disposed radially inwardly of said inner shell; said inner shell including a flow reεiεtant porous structure for at least a part of its axial extension to provide for permeation therethrough of said treatment solution and to present flow resistance to said residual of said liquid mixture substantially uniformly along the outer surface of εaid porouε εtructure; a liquid mixture inlet to one end of εaid gap; a treatment εolution inlet for introducing εaid treatment εolution into the radially outward portion of εaid gap; a frame rotatably εupporting said rotary assembly for rotation about said axis; a drive coupled with εaid rotary assembly to rotate said assembly at a predetermined fraction speed sufficient to, when said mixture is introduced to said gap through said mixture inlet and said treatment solution is introduced through said solution inlet, apply a settling velocity to said predetermined fraction sufficient to hold it suspended within said gap while said treatment solution is introduced through said treatment solution inlet causing it to flow radially inwardly through said suspension liquid and second fraction.

31. The apparatus of Claim 30 for use in washing a mixture of the type including red blood cells with a saline wash solution and that includes: a saline tank encapsulating said rotary assembly and formed with said treatment solution inlet port; and a treatment pump to pump εaid εaline treatment solution into said tank and through said treatment solution inlet port.

32. The apparatus of Claim 30 wherein: said inner shell is formed adjacent said liquid mixture inlet with a frusto-conical shape to slope gradually radially inwardly and away from said fluid mixture inlet to gradually increase the radial thickness of εaid gap.

33. The apparatuε of Claim 30 that includes: bearings interposed between said housing and assembly; and circulation means connected with said treatment εolution inlet for flowing εaid treatment solution in heat exchange relationship with said bearings.

34. Fractionating apparatus for fractionating a first fraction suspended in a suspension medium of a liquid mixture including a second fraction, said fractions formed by respective components having settling velocities which are, under identical conditions of centrifugal force, higher and lower, respectively, than a selected range of settling velocities, said apparatus comprising: an axially elongated rotary assembly adapted for rotation about a central axis, said assembly having inner and outer shells defining therebetween an axially extending tubular gap and an inner region disposed radially inwardly of said inner shell; said inner shell including a flow resistant porous εtructure for at least a part of its axial extension to provide for permeation therethrough of said first fraction and to present flow resistance to said residual of said liquid mixture substantially uniformly along the outer surface of said porous structure; a liquid mixture inlet to one end of said gap; a frame rotatably supporting said rotary assembly for rotation about said axis; and a drive coupled with said rotary assembly to rotate said asεembly at a predetermined fractioning εpeed εufficient to, when εaid mixture iε introduced to εaid gap through εaid mixture inlet, apply εettling velocitieε in said selected range of settling velocities to hold said second fraction in said gap while flowing said firεt fraction through εaid flow reεiεtant porouε εtructure into εaid inner region.

STATEMENT UNDER ARTICLE 19

Claims 1, 2, 15, 18, 24, 27, 28, 29, 30, 32 and 34 have been amended to more particularly claim the matter for which protection iε sought and are being submitted in accordance with PCT Article 19.

The amended claims are fully supported by the description and do not go beyond the disclosure in the international application as filed. None of the references in the International Search Report, alone or in combination, suggest the present invention as now claimed.

In conclusion, the amended claims more specifically claim the present invention disclosed in the international application. The present invention as now specifically claimed is not suggested by the international search authority references, alone or in combination.