US 3380584 A
Description (OCR text may contain errors)
April 30, 1968 M. J. FULWYLER PARTICLE SEPARATOR 5 Sheets-Sheet 1 Filed June 4, 1965 ACOUSTIC DRIVER ACOUSTIC GENERATOR DRIVER 33 DELAY PROPORTIONAL PULSE GENERATOR COLLAR PULSE GENERATOR GENERATOR E s l- U P o T INVENTOR. Mack J. F u/wy/er April 30, 1968 M. .1. FULWYLER PARTICLE SEPARATOR 5 Sheets-Sheet 2 Filed June 4, 1965 PHOTO-CELL ASSEMBLY Fig.
INVENTOR. Mack J. Fu/wy/er April 30, 1968 M. J. FULWYLER PARTICLE SEPARATOR 5 Sheets$heet Filed June 4, 1965 wxON v: 000+ mm. E5260 3332735565 a 952% tiSo IN V EN TOR. Mack a! F u/wy/er United States Patent 3,380,584 PARTICLE SEPARATUR Mack .l'. Fulwyler, Los Alamos, N. Mex., assiguor to the United States of America as represented by the United States Atomic Energy Commission Filed June 4, 1965, Ser. No. 461,566 3 Claims. (Cl. 209--3) This invention relates to an apparatus and method for sorting minute particles present in a suspension fluid in accordance with a selected characteristic. The selected characteristic may be size, presence of radioactivity, color, fluorescence, light transparency or opaqueness or in fact any characteristic which can be translated into an analogue electrical quantity.
An apparatus capable of sorting minute particles in accordance with distinctive characteristics among the particles themselves has many important uses. For example, in the field of biology, it would be useful to sort blood cells in accordance with size. In investigations of carcinoma, it is useful to separate cancer cells from normal cells as indicated by size, radioactivity or luminescence from normal cells. In the separation of rare metals from ores, it would be useful to comminute the material and separate reflective metal particles from the refuse on the basis of reflectivity translated by electrooptic means into analogue electrical quantities.
In the present state of the art, means are available for counting minute particles in a fluid but no apparatus or method is available for physically sorting and counting such particles in accordance with the magnitude of a selected characteristic.
It is accordingly a principal object of the present invention to provide a method and apparatus for detecting minute particles in a suspension fluid and to physically separate such particles into groupings of particles having a common characteristic.
It is a further object of my invention to provide improved apparatus for separating a moving stream of fluid containing suspended particles into separate droplets and to deflect such droplets into separate receptacles each designated to catch and retain droplets containing particles of a predetermined characteristic.
The above and additional objects and advantages of the present invention will be made evident as the specification proceeds with reference to the following specification and claims and the illustrations in the accompanying drawings.
Referring to the drawings:
FIGURE 1 is a schematic diagram of a simplified system of the present invention;
FIGURE 2 is a pictorial illustration of a nozzle assembly and deflection system;
FIGURE 3 is a detailed cross-sectional view of a preferred nozzle adapted to develop the jet and in response to each particle passing through the nozzle to generate an analogue electrical quantity;
FIGURE 4 is a cross-sectional view of the nozzle adapted to generate analogue electrical quantities in response to luminescence of particles;
FIGURE 5 is a cross-sectional view of the nozzle adapted to generate electrical analogue quantities in response to radioactivity of particles; and
FIGURES 6, 7 and 8 are an electrical schematic diagram of a control system for a working embodiment of the invention.
The present invention in its broad sense is a combination of means for converting a particle character istic into an electrical analogue quantity, and means for deflecting the particle in accordance with the electrical quantity into a corresponding repository. For example,
the separation of biological cells can be accomplished by combining, in accordance with the teachings of this invention, the teachings of W. H. Coulter in Patent 2,656,508, issued Oct. 20, 1953, with the teachings of Richard G. Sweet in Report SEL64-OO4 of the Stanford Electronics Laboratories, of Stanford University, published March 1964 and also identified as Technical Report No. 1722-1 prepared under US. Signal Corps contract.
The teachings of Coulter Patent 2,656,508 of interest to the present invention is the provision of a short dielectric duct or aperture of appropriate cross section through which the suspension fluid is forced to flow, and a pair of electrodes disposed on opposite ends of the aperture in contact with the fluid. The electrical resistance between the electrodes is aflected by the presence and size of a particle in the fluid within the duct.
The teachings of Sweet (SEL-64-004) applicable to the present invention are the provision of an orifice through which the fluid is caused to flow and the application of vibration to the orifice which causes the fluid jet issuing therefrom to separate into discrete, equal volume droplets. Electrostatic droplet charging means are associated with the jet close to and downstream from the orifice where the jet separates into droplets to impose an electric charge on each droplet. Fixed potential electrostatic deflecting means downstream of the droplet charging means causes each droplet to deflect an amount related to the charge on the droplet.
In FIGURE 1 I have illustrated, in diagrammatic and block form,t'ne combination of elements which form one practical embodiment of my invention. Liquid containing particles is suspension are stored in any suitable container 11. Means for pressurizing the fluid may be any suitable method such as a pump in the outlet of container .11, or gas pressure means 13 and controllable pressure reducing means 15.
The fluid container 11 is connected by a strainer 17 and pipe 19 to nozzle 21. Strainer 17 is provided to pass particles within the range of interest and to stop particles of gross size from clogging the ystem. Although the fluid in passing through the nozzle may be jetted and separated into droplets by suitably vibrating the nozzle, I prefer to apply pulsations directly to the fluid and avoid vibration of the nozzle. I accomplish this method of drop formation by resilient acoustic insulation 20 between nozzle 21 and acoustic coupler 23. The acoustic coupler 23 is directly driven by an electrically driven vibrator such as piezoelectric driver 25 which in turn is energized by adjustable frequency generator 27.
The frequency and amplitude of the output of generator 27 is correlated with the viscosity and velocity of fluid passing through the nozzle in order to produce equal size discrete droplets from the resulting jet.
Referring to FIGURES 1 and 2, sensor element 29 is shown associated with the fluid-containing system, preferably at the outlet of the nozzle. At the outlet of the nozzle the fluid has obtained its je-t cross section and velocity so that the transport time between the passage of the fluid through the sensor and the point along the path of break-up of the jet into droplets is a fixed and calculable quantity.
The sensor 29 is selected to be responsive to the particular characteristic of the particles of interest, that is, it will be responsive to particle size, radioactivity, fluorescence, luminescence, electrical conductivity, light transmissibili-ty, light reflectivity of any other characteristic which is capable of transformation into an electrical quantity.
The electrical pulse generated in the sensor by the passage of a particle is amplified and shaped by proportional pulse generator 31. The utilization of the pulse to deflect the associated droplet is delayed an amount of time equal to the time consumed by the fluid containing the particle in traversing the distance from the sensor to the point Where the liquid is detaching from the 'jet as a droplet. The pulse after appropriate delay in delay element 33 controls the amplitude and time of generation of droplet charging potentials. More specifically, each droplet or a group of droplets is charged by collar pulse generator in response to the delayed pulse from delay element 33. The collar pulse generator 35 produces the proper potential which is inductively impressed along the jet stream between charging collar 37 and sensor 29. The charging collar surrounds the jet stream at a location which includes the droplet separation zone. Thus, the potential which exists between the collar and the sensor at the time of separation of the droplet, determines the bound charge on the droplet. The droplet, or train of droplets, continues in the direction of the jet stream until it passes between deflecting plates 39 and 41. Deflecting plates 39 and 41 are energized with a steady state potential of suitable magnitude. Each droplet in passing between the deflection plates is deflected an amount determined by the charge on the droplet. The droplets are thus segregated in accordance with the amount of a selected characteristic and can be collected in any desirable manner. FIGURE 2 shows collection in ten catch basins but in some cases it is desirable to collect the segregated droplets on a moving strip of blotting paper. The element 43 in FIGURES 1 and 2 is intended to illustrate in general such various types of collection systems.
The system of FIGURE 1 is a simplified system illustrated to show generally the manner of accomplishment of the objectives of this invention.
The particle responsive sensor 29 is selected to generate an electrical quantity in response to a selected characteristic. Where sorting and collection in accordance with the presence of and size of particles is desired, the sensor preferably takes on the form of FIGURE 3. The sensor comprises the metallic end of nozzle 21, a plate of dielectric material 22 and an end electrode 24. Particles passing through the aperture in dielectric plate 22 cause a change in electrical resistance between nozzle body 21 and end electrode 24. This change in resistance is transformed into an electrical pulse in proportional pulse generator 31 as described with reference to FIGURE 1.
In the separation of blood cells, a typical set of conditions is as follows:
The concentration of the blood cells in the saline solution is preferably maintained so that not more than one cell will appear in several, i.e., for example, seven droplets. The collar pulse is of a duration to impress an identical charge on each droplet in the sequence of seven. Although it would be preferable to treat each droplet separately, thus enabling a faster separation of particles from a more concentrated solution, the velocity across the cross section of the nozzle is not so uniform as to result in complete assurance that a detected particle will always appear in a certain droplet instead of the next droplet following. In the interest of reliability in sorting, a dilute solution is preferred and the collar charging pulse preferably brackets one or more droplets ahead and behind the droplet expected to contain the particle.
As is shown in FIGURE 3, the mass of the nozzle is Nozzle dielectric 22 aperture "microns" 42 Electrode 24 aperture do 36 Frequency of acoustic driver kilocycles 72 Fluid (saline solution) percent 0.9 Fluid pressure p.s.i 53 Transit time sensor 29 to collar 37 microseconds 200 Duration of charging pulse on collar do 200 Velocity of jet meters/sec 14.6 Size of human red cell cubic microns" 85 Size of mouse red cell do 42 eliminated from the vibratory system by acoustic insulators 45-47. The effect of this is a reduction in demand on the acoustic driver and considerably greater effectiveness of the vibratory energy in breaking up the jet into uniform droplets. In addition, interference with clean application of the acoustic vibration to the fluid by the inertia of the nozzle mass and consequent multiple moding is avoided.
The improved nozzle of the present invention comprises a nozzle proper 21, preferably of platinum. Acoustic coupling rod 23 is resiliently isolated from both the nozzle 21 and the housing 49 of the nozzle assembly by resilient O-rings 45 and 47. The nozzle housing 49 includes the synthetic plastic shell 51 which contains a cylindrical bore in which the nozzle body 21 is preferably a slight squeeze fit. The nozzle is anchored in shell 51 by any convenient means such as cast-in-situ synthetic resin 53. Both the nozzle and the shell 51 are provided with annular grooves to permit resin 53 to permanently lock the two in assembled relation.
The acoustic driver rod at its end portion has a diameter equal substantially to the largest diameter of the tapered nozzle 21. The diameter of the driver end portion is reduced from that of the remainder of the driver rod for a distance adequate to accommodate O-ring 47. The end portion of the drive rod is thus hydraulically sealed to the large end of nozzle 21 by O-ring 47. The drive rod 23 is also provided with an integral enlarged portion 52 in order to provide a shoulder for O-ring 45 to bear against. Annular collar 54 provides a shoulder 55 for holding the assembly together with O-rings 45 and 47 under appropriate pressure when upper body member 57 is threaded onto body member 51. Fluid duct 19 is inserted into nozzle 21 with a hermetically tight fit.
It follows that acoustic drive rod 23 is substantially afree piston as far as any part of the nozzle or nozzle assembly is concerned.
As mentioned above, although particle separation by size is important, other types of sensing are also possible and important. FIGURE 4 shows one form of sensor which is luminescent responsive. A light source 60 capable of emitting light of the selected frequency is coupled to the fluid duct by transparent window 61. Particles which luminesce will emit light shortly after excited and will be detected by photocell detector 63. The output of detector 63 is utilized as indicated in FIGURE 1.
FIGURE 5 schematically shows a sensor 29 to detect radioactive particles such as thyroid cancer cells. A scintillation element 69 is supported in cylinder 67 and in the path of pick-up of photo-sensitive assembly 63. Numerous other types of sensors may be used, the only requirement being that the particle characteristic of interest be translatable into an electrical quantity by any of the methods known in the art.
Referring to FIGURES 6, 7 and 8 a complete electronic control system for association with the particle sensor and droplet charging system is shown. This system amplifies the pulses generated by particles passing through the sensor 29, converts the pulse height information into digital form, stores the digital information in storage flipfiops and after a time delay equal to the transit time of the liquid in passing from the sensor to the charging collar, develops a pulse having a width tailored to the number of droplets in each group and the amplitude of which is determined by the stored digital information.
The electronic system contains 1) numerical logics which is concerned with the conversion of the digital information from analogue-to-digital converter 101 into a charging collar potential of corresponding magnitude, and (2) control logics which is concerned with the sequence of operation of the numerical information system.
The description of the electronic system to follow below will be more easily understood if the requirements of the system are first outlined.
The system is not to respond to particles having the characteristic of interest below :a selected level. This is accomplished by lower boundary, or threshold discriminator 118.
All particles or multiple particles causing a sensor response in excess of a selected range are to be segregated in a maximum throw collector. This is accomplished by upper boundary discriminator 119. An output from this discriminator results in a maximum available collar charging potential in order to kick such particles into an end basin.
There is a certain time lapse between the time a particle causes a sensor signal and the time at which the same particle in a droplet separates from the jet. This time lapse produces two requirements, one of which is the storage of the digital output of the analogue-to-digital converter 101 until the information can be used, and the second requirement is the provision of an accurate time lapse control. The first requirement is met by an information storage system, which in this selected embodiment are storage flip-flops 103. The second requirement is met by the provision of charging delay univibrator 157 (FIG. 7
The collar charging potential must be the correct value corresponding to the magnitude of the particle characteristic of interest. This is accomplished by processing the output of analoguc-to-digital converter 101 through storage flip-flops 103, through binary-coded-decimal to decimal converter 129, through charging pulse forming network 132 and digital-to-analogue converter 135.
The charging collar potential for each droplet containing a particle must exit long enough for a droplet to be charged. This requires that the pulse be a rectangular pulse of correct length. This duration is obtained by charging time univibrator 159 which triggers-on for the correct duration the appropriate and circuit in the pulse forming network 132. This in turn results in the correct rectangular pulse at terminal 147 (FIGURE 8) in the emitter-follower circuit of transistor 143.
The output of sensor 29 is fed into pulse generator 31 and from there to the input of analogue-to-digital converter 101. The capacity of the analogue-to-digital converter (multi-channel analyzer) is chosen to accord with the selected number of separations desired. Each digit is encoded into a 1-2-4-8 decimal code. In the embodiment herein described, the particles are to be collected in ten catch basins 43, and because for purposes of convenience, an analogue-to-digital converter of 100 channels was used, only the tens of the address are used.
The tens pulses of the address are current amplified in emitter followers 104-107 and pass through and gates 111-117 Which are inhibited if the pulse from the sensor 29 is greater than the selected upper limit set by upper discriminator 119. However, a pulse from upper discriminator 119 is used to deflect an oversize particle into the ten basin (see FIGURE 2). To this end, the upper discriminator pulse is shaped in Schmitt trigger 144, generates a negative pulse of approximately 85 microseconds in duration in univibrator 127 which inhibits and circuits 111-117 but also directly activates or circuits 121 and 123 which inserts the 8 and 2 bits into storage flip-flops 103-B and 103-D through and circuits 125-B and 125-D.
The upper discriminator pulse also passes through or circuit 128 and after a 5 microsecond delay generates a 5 microsecond transfer pulse which inserts the ten into the storage flip-flops which, as will be seen later, results inthe application of a voltage to the charging collar of magnitude in excess of that selected for segregation of particles in the lower ranges of interest.
If the sensor 29 has detected a particle which, for example, is to be collected in the third catch basin of FIG- URE 2, two signals are originated. The lower discriminator 118 delivers an activation pulse which travels along line A to Schmitt trigger 148 (FIGURE 7). The same signal is converted in analogue-to-digital converter 101 into a "3 which is a positive output on the 10 and 20" addresses. These ones pass through emitter followers 104 and 105 to and gates 111 and 113. These and gates are not inhibited by the presence of an upper bound pulse and therefore insert ones through and gates A and 125-3 into the storage system 103. The analogueto-digital converter upon completion of a count also gencrates an add-l pulse. This pulse triggers Schmitt trigger 155 which delivers a pulse through or circuit 128 to transfer delay circuit 130 which after a 5 microsecond delay triggers transfer-pulse-generator 131 which generates a 5 microsecond long positive square transfer pu se.
The occurrence of the transfer pulse on and circuits 125-A to 125-D, together with ones from the analogue-to-digital converter on 125-A and 125-3, causes a positive pulse to appear on the output of the flip-flops 103-A and 103-B. The flip-flops and emitter followers in the output of the flip-flops arearranged so that a 1 is negative and a 0 is positive. Therefore, a "1 on the input of flip-flops 103-A and 103-B and a 0 on the input of flip-flops 103-C and 103-D gives rise to the following conditions on terminals E to L: a negative voltage on F, H, I and K and positive voltages on E, G, J and L. There are, therefore, plus voltages on buses 1, 2, 4 and 8 of the binary-coded-dccimal to decimal converter 129. These simultaneous positive voltages cause a plus output from and circuit 131-C. The plus potential remains on the output of the and circuit 131-0 until an enabling signal is put on line D.
The generation of the enabling pass signal on line D takes place in the apparatus of FIGURE 7. The signal pulse from the lower discriminator passes through normally enabled and gate 153. This pulse triggers charging delay generator 157. Charging delay generator 157 delivers a block-out triggering pulse 154 to cause blockout or gate 155 to start the block-out pulse. At the end of the delay time set in charging delay generator 157 a triggering pulse is generated to trigger charging time generator 159. The output of charging time generator 159 also goes into the input of or circuit 155 so that a blockout or inhibiting potential is applied to and gate 153 from just after time to start signal on line A is received until the corresponding droplet is separated from the jet. The output pulse from the charging time generator, together with the add-1 pulse, passes through and circuit 152 and alters the state of flip-flop 154. The output of flip-flop 154 passes a pulse onto line D when and gate 161 is enabled by the output pulse from charging time generator 159, and only during such time of enabling. The next pulse from the lower discriminator resets flipflop 154.
The occurrence of the steady state potential on the D line (FIG. 8) enables the appropriate and gate in the charging pulse formation network 132 to pass the output of the activated one of gates 131-A to 131-J to the digital-to-analogue converter 135. In the example herein treated. and gate 131-C is enabled by the potential on line D to pass a pulse through its emitter follower to input (3) which open-circuits transistor -C. This permits the 90 volt potential on the collector of transistor 135-C to raise the base potential of output transistor 143. The resistor string 149-A to 149-K is proportioned so that the output voltage on terminal 147 has a value of 90 volts.
It was stated supra that an outsize particle resulted in a maximum collar potential at terminal 147. The signal from upper discriminator 119, after processing by Schmitt trigger 144 (FIG. 6) and in univi'orator 127 generates an inhibt potential on and circuits 111 to 117 and goes through or circuits 121 and 123. Therefore, flip-flops 103-B and 103-D are switched. Therefore, negative potentials appear on terminals E, H, I and L and positive potentials appear on F, G, J and K. Referring to FIG- 7 URE 8 it is seen that buses I, Z, 2 and 8 of the binarycoded-decimal to decimal converter carry positive potentials and that and circuit 131] is the only and circuit enabled. This in turn triggers-off transistor 135-] and results in 300 volts approximately on terminal 147.
The construction and operation of the digital-to-analogue converter 135 is now described. Each transistor 135A to 135-] is connected through its respective collector resistor 134-A to 134-] to a 600 volt potential source through a string of voltage dividing resistors 149A to 1494. The values of the dropping resistors were chosen in this example to provide supply voltages to the cathodes of diodes -137-A to 137-] which increase by 30 volt increments. Thus a decoded 1 goes to transistor 135A with a 30 volt supply, and a decoded 10 goes to transistor 135-] with a 300 volt supply.
With no inputs, all of transistors 135A to 135] are on and in a saturated condition so that the collectors are at less than one Volt above ground. When a negative pulse is received it applies a reverse bias to the baseemitter junction of the corresponding transistor which turns the transistor ofi. As the collector current falls to zero, the voltage on the collector charges up to the voltage on the supply voltage resistor divider string and passes through the respective diode (137A137-J) to the base of emitter follower 143. The voltage pulse appearing on the emitter of emitter follower 143 is applied to charging collar 37.
The particular electronic system above described is presented only for purposes of illustrating a complete embodiment and not in any limiting sense. Any other electronic system capable of performing the desired functions may be used.
What is claimed is:
1. Apparatus for separating small particles suspended in a fluid in accordance with the amount of a selected characteristic of each particle comprising means for jetting the fluid, sensor means surrounding the path of the fluid jet for generating an analogue electrical quantity in response to the passage of a particle, means for acoustically pulsing the fluid jet to cause separation into droplets in a discrete separation zone, electrical charging means surrounding the jet path at the separation zone, electrical delay and charging means connected between said sensor and said electrical charging means having a period of delay equal to the time of passage of a particle passing from the sensor means to the droplet separation zone, electrical deflecting means supported symmetrically about the path of the droplets, whereby each droplet containing a particle is deflected an amount corresponding to the amount of characteristic of the particle contained therein.
2. A device for sorting minute particles suspended in a fluid comprising a nozzle for jetting the fluid, a particle sensor supported adjacent the outlet of the nozzle for generating an electrical quantity in response to a selected characteristic of a passing particle, means for acoustically pulsing the fluid jet in the direction of its motion whereby the jet separates into droplets in a fixed zone, a conducting charging collar surrounding the jet and supported to surround the zone of droplet separation, electrical charging means connected through electrical delay means to the particle sensor to develop a collar charging potential having a magnitude related to the magnitude of output of the sensor, said electrical delay means electrically connected between the sensor and the electrical charging Cir means having a delay period equal to the transport lapse time of fluid passing from the sensor to the zone of drop let separation whereby each droplet containing a particle is impressed with a bound charge having a magnitude related to the magnitude of the characteristic of the particle in the droplet, a pair of electrical deflecting plates supported symmetrically about the train of droplets down stream of the charging collar whereby each droplet is deflected an amount proportional to the amount of characteristic of the particle contained in the droplet, and catch basins support in a plane normal to the jet path for collecting the segregated particles.
3. Apparatus for sorting minute particles suspended in a fluid comprising, a nozzle for jetting the fluid, means supported on the nozzle for acoustically pulsing the fluid passing therethrough, a particle sensor consisting of a pair of electrodes spaced by a dielectric separator supported on the outlet end of said nozzle, said electrodes and said dielectric separator each containing a central aperture having a diameter adapted to pass with slight clearance the largest particle of interest and said electrodes and dielectric separator being supported with their apertures aligned with one another, means for impressing a potential across said electrodes whereby each particle passing through the apertures generates an electrical quantity, said electrodes being connected to a multi-channel analogue-to-digital converter having digital outputs and an add-1 pulse output, electrical memory storage means connected to the digital output of the analogue-todigital converter, electrical delay means having a selected time constant connected between the analogue-to-digital converter add-1 output and the memory storage means for transferring to output terminals the content of the storage means at the end of the selected time constant, means electrically connecting the output of the memory storage means to a binary-coded-decimal-todecimal converter and means electrically coupling the output of the binarycoded-decimal-to-decimal converter to a digital-to-analogue converter whereby decimally related potentials are generated in the output of the digital-to-analogue converter corresponding to ranges of particle sizes, a cylindrical metal charging collar supported colinearly with the nozzle at the zone of separation of the jet into droplets, said electrical delay means selected time constant being equal substantially to the duration of time consumed by fluid in passing from the sensor to the droplet separation zone, said charging collar being electrically connected to the output of the digital-to-analogue converter, electrical charge deflecting means supported symmetrically about the axis of the nozzle and down stream of the charging collar and a plurality of droplet collection basins supported below the deflection means and along a path normal to the nozzle axis.
References Cited UNITED STATES PATENTS 2,646,880 7/1953 Frankel 20 91 1 1.6 2,656,508 10/ 1953 Coulter 32471 2,756,388 7/1956 McLean 324--32 3,123,541 3/1964 Donnell. 3,124,172 3/ 1964 Paxson.
FRANK W. LUTTER, Primary Examiner.