HK1164406B - Magnetically driven micropump - Google Patents

Description

Magnetically driven micropump

Cross Reference to Related Applications

This application claims priority from U.S. provisional patent application No.61/152,165 filed on 12.2.2009, which is incorporated herein by reference in its entirety.

Background

The present disclosure relates to a magnetically driven micropump for handling small fluid volumes. In particular, the present disclosure relates to micropumps that include a magnetically actuated membrane to transfer a fluid.

The field of microfluidics generally involves handling very small fluid volumes on the order of a few nanometers. Microfluidic technology has increasingly important applications in fields such as life sciences and chemical analysis. Microfluidic technology devices, also known as micro-mechanical systems (MEMS), include devices for fluid control, fluid measurement, medical testing, DNA and protein analysis, active drug delivery, and other biochemical applications.

Typical fluid flow rates for micropumps range from about 0.1 microliters per minute to a few (80-180) milliliters per minute. Flow rates on this scale are useful in applications such as disposable micro total analysis systems (μ TAS) or lab-on-a-chip (LOC) for chemical and biological analysis, point of care testing for medical diagnostic testing, implantable drug delivery systems requiring fine-grained management and accurate control for drug delivery (e.g., insulin), and cardiology systems for blood transfusion and pressurization.

Since most MEMS processing technology originated from microelectronics, the first silicon micropump in the 1980 s was based on piezoelectric actuation of a thin membrane, primarily for controlled insulin delivery systems. This work illustrates the feasibility of silicon-based micropumps and the creative profound research on silicon micropumps. Moreover, in the pharmaceutical and clinical therapeutic fields, a number of commercial implantable silicon micropumps are reported for insulin delivery and therapeutic agent administration.

Recently, a variety of polymeric materials and novel microfabrication techniques, such as soft lithography, micro-lithography, micromolding and polymeric surface micromachining, have been studied and developed for ever-increasing low-cost, integrated and miniaturized disposable μ TAS applications. Many polymeric materials, including plastics and elastomers, have been increasingly incorporated into other microdevices as substrates, structural members, and functional members due to their excellent mechanical properties, good chemical resistance, and low manufacturing cost. Among the most commonly used polymers, Polydimethylsiloxane (PDMS) has been widely used in microfluidic devices due to excellent biocompatibility, simple manufacturing processes (molding and reversible bonding), and optical transparency (ease of monitoring and interrogation) and elasticity (good sealing and connectivity).

Silicon-based and plastic-based valveless micropumps were used as examples for comparison with polymer-based micropumps. The fabrication process of silicon-based micropumps includes three sequential deep reactive particle etching (DRIE) steps and one silicon-glass anodic bonding step, while LIGA, micro-injection, or hot embossing molding and assembly of multiple sheets by adhesive or bolts are involved for plastic pumps. On the other hand, for PDMS based micropumps, only a multi-layer soft lithography process and PDMS-PDMS bonding technology is required. From a manufacturing cost perspective, PDMS-based micropumps are much lower than the first two types of micropumps.

Moreover, a major challenge of plastic micropumps is high fluid leakage due to the surface roughness of the thin plastic layer. Bolting makes matters worse because stress is concentrated at the bolted locations on the interface between the layers. Adhesive bonding also tends to promote clogging of the microstructures. Thus, PDMS is a practical material for micropumps (short processing time and low cost).

Disclosure of Invention

The operating principle of the micropump disclosed herein is that the vibrating diaphragm causes a pressure change in the chamber, which directs the dynamic flow of the fluid tube in the form of a passive valve. Typically the passive valves are incorporated as check valves in the inlet and outlet of a reciprocating micro-pump in the form of a cantilever flap, a bridging diaphragm, a round ball, a moving structure, a nozzle/diffuser or a tesla element. However, the valveless micropump integrating the nozzle/diffuser element has particular benefits for disposable μ TAS applications, for example in biomedicine and biochemistry, since the risk of suspended particles clogging, wearing and fatigue moving mechanical parts can be reduced and virtually eliminated. Moreover, the planar nature and simple implementation of the nozzle/diffuser results in low cost and miniaturization of the micro-pump for disposable applications.

The valveless micropump of the present disclosure includes a nozzle and diffuser element, a fluid chamber, and a vibration-actuated membrane. The diaphragm is integrated with a small block magnet, which has the advantage of a large attractive or repulsive magnetic force and deflection of the diaphragm. The alternating vertical magnetic force on the diaphragm results in a large volume stroke, which is desirable for high flow rate micropumps. Furthermore, magnetic actuation is an externally applied field, in which case the micro-pump is controlled by an air gap. Thus, electrical connectors for applying current or voltage on the micro-pump can be avoided, which also provides the possibility of miniaturization in μ TAS applications.

The principles and operation of the subject matter of the present disclosure are explained fully in the Zhou et al article (Fluid DampingEffects on resonance Frequency of an electromagnetic-affected valve microprocessor, International Journal of Advanced Manufacturing Technology, April 24, 2009), which is incorporated herein by reference in its entirety.

One aspect of the present disclosure includes a micropump for delivering a fluid. The micropump includes a pump assembly having a first pump body defining a first fluid flow path. The first pump body includes a first chamber including a first chamber wall and a first sidewall; a first inlet and a first outlet, wherein the first inlet and the first outlet are in fluid communication with the first chamber. The pump assembly also includes a second pump body defining a second fluid flow path. The second pump body includes a second chamber including a second chamber wall and a second side wall; a second inlet and a second outlet, wherein the second inlet and the second outlet are in fluid communication with the second chamber. The pump assembly further includes a flexible diaphragm disposed between the first chamber and the second chamber. The micropump further includes an actuator assembly configured to cooperate with the pump assembly. The actuator assembly includes a driver magnetically coupled to the diaphragm, and a sensor configured to detect a position of the diaphragm, wherein the driver applies a magnetic force to the diaphragm causing the diaphragm to deflect, and wherein the deflection of the diaphragm causes a change in pressure within the first and second chambers, thereby causing fluid flow.

Another aspect of the present disclosure includes a micropump assembly for delivering a fluid from a fluid reservoir, the micropump assembly including a pump barrel. The pump barrel includes a first pump body defining a first chamber including a first chamber wall and a first sidewall; a first inlet and a first outlet, wherein the first inlet and the first outlet are in fluid communication with the first chamber. The pump body further includes a second pump housing defining a second chamber including a second chamber wall and a second side wall; a second inlet and a second outlet, wherein the second inlet and the second outlet are in fluid communication with the second chamber, and a flexible diaphragm disposed between the first chamber and the second chamber, wherein the pump barrel is configured to allow fluid communication from the fluid reservoir to at least one of the first chamber and the second chamber. The micropump assembly further includes a housing enclosing an actuator assembly configured to cooperate with the micropump cartridge. The actuator assembly includes a driver magnetically coupled to the diaphragm, and a first sensor configured to detect a position of the diaphragm, wherein the driver applies a magnetic force to the diaphragm causing the diaphragm to deflect, and wherein the deflection of the diaphragm causes a change in pressure within the first chamber and the second chamber, thereby causing fluid flow. The micropump assembly further includes a controller coupled to the driver and configured to control the diaphragm position by receiving a signal from the first sensor and adjusting the magnetic force applied by the driver. The micropump assembly further includes a power source configured to energize the driver and the controller, wherein the housing is configured such that the micropump cartridge can be inserted and retained in the actuator assembly.

Another aspect of the present disclosure includes a method of manufacturing a micropump. The method comprises the following steps: manufacturing a flexible diaphragm from a polymer material, comprising the steps of: spin coating a first polymer layer on a silicon wafer and allowing the first polymer layer to cure, disposing a magnetic material on the first polymer layer, applying a second polymer layer around the magnetic material and allowing the second polymer layer to cure, and applying a third polymer layer and allowing the third polymer layer to cure; manufacturing a rigid pump body by pouring a liquid polymer material into a mold configured to form the fluid chamber, the inlet channel, and the outlet channel, and to allow the liquid polymer to solidify; aligning the flexible diaphragm with the rigid pump body; and bonding the flexible polymer diaphragm to the rigid pump body.

Another aspect of the present disclosure is a micropump for delivering a fluid. The micropump includes a pump assembly having a first pump body defining a first chamber. The first chamber includes a first chamber wall and a first sidewall, a first inlet and a first outlet, wherein the first inlet and the first outlet are in fluid communication with the first chamber, and a first flexible diaphragm disposed on the first chamber opposite the first chamber wall. The pump assembly has a second pump body defining a second chamber including a second chamber wall and a second sidewall, a second inlet and a second outlet, wherein the second inlet and the second outlet are in fluid communication with the second chamber, and a second flexible diaphragm disposed on the second chamber opposite the second chamber wall. The pump assembly further comprises at least a third pump body arranged between the first and second pump bodies. The third pump body defines a third chamber including a third sidewall, a third inlet, and a third outlet, wherein the third inlet and the third outlet are in fluid communication with the third chamber, wherein the at least third chamber is adjacent to the second diaphragm and the second diaphragm. The micropump further includes an actuator assembly configured to cooperate with the pump assembly. The actuator assembly includes a driver magnetically coupled to the first and second diaphragms, and at least one sensor configured to detect a position of the first and second diaphragms, wherein the driver applies a magnetic force to the first and second diaphragms causing the first and second diaphragms to deflect, and wherein the deflection of the first and second diaphragms causes a change in pressure within the first, second, and third chambers, thereby causing a fluid flow.

Drawings

The disclosure will be described below with reference to the accompanying drawings, given only by way of non-limiting example, in which:

FIG. 1 is a perspective view of an embodiment of a micropump assembly of the present disclosure;

FIG. 2 is an exploded view of the micropump assembly of FIG. 1;

FIG. 3 is a perspective view of a pump body having a nozzle/diffuser flow element for generating unidirectional flow;

FIG. 4 is a schematic view of a nozzle/diffuser flow element having a frustoconical configuration;

FIG. 5 is a schematic view of a nozzle/diffuser flow element having a frusto-pyramidal configuration;

FIG. 6 is a schematic diagram of a single chamber micropump illustrating a fluid flow path;

FIG. 7 is a cross-section of the dual chamber micropump assembly of the present disclosure showing a combination of parallel flow paths;

FIG. 8 is a perspective view of an embodiment of a micropump assembly of the present disclosure;

FIG. 9 is an exploded perspective view of the micropump assembly of FIG. 8;

FIG. 10 is a perspective view of a micro pump cartridge and actuator assembly receptacle;

FIGS. 11 through 14 are diagrams of finite element models of embodiments of diaphragms of micropumps of the present disclosure;

FIG. 15 is a graph of the resonant frequency of a valveless micropump used in embodiments of the present disclosure as a function of diffuser slenderness ratio;

FIG. 16 is a graph of the resonant frequency as a function of diffuser opening angle for an embodiment of a valveless micropump of the present disclosure;

FIG. 17 is a graph of resonance frequency as a function of diffuser aspect ratio for an embodiment of a valveless micropump of the present disclosure;

FIG. 18 is a graph of the variation of the resonant frequency with the thickness ratio of cavity depth and diaphragm thickness for an embodiment of a valveless micropump of the present disclosure;

FIG. 19 is a graph of diaphragm displacement versus time for an embodiment of a micro-pump of the present disclosure;

FIG. 20 is a finite element model of an exemplary embodiment of the valveless micropump of the present disclosure illustrating maximum displacement in the discharge mode;

FIG. 21 is a finite element model of an exemplary embodiment of the valveless micropump of the present disclosure illustrating maximum displacement in the pumping mode;

FIG. 22 shows a microfluidic connector for use with the micropump of the present disclosure;

FIG. 23 is a graph showing maximum pumping flow rate as a function of excitation frequency at different actuation currents;

FIG. 24 is a graph of maximum flow rate at different actuation current amplitudes;

FIG. 25 is a comparison of maximum pumping flow rates;

FIG. 25A is a graph of a square wave excitation signal before and after actuator loading;

FIGS. 26 to 28 are graphs showing flow rate as a function of excitation frequency;

FIG. 29 is a graph of actuator temperature over time;

FIGS. 30-32 are schematic views of an embodiment of a micropump of the present disclosure during operation;

FIG. 33 is a schematic diagram of an arrangement of magnetic position sensors;

FIGS. 34 and 35 are schematic diagrams of the magnetic field strength of the actuator coil and magnet, respectively;

FIG. 36 is a graph of magnetic field perturbations;

FIG. 37 is a graph of magnetic field strength;

FIG. 38 is a graph of magnetic field perturbations measured by a sensor;

FIG. 39 is a graph of magnetic field strength measured by a sensor;

FIG. 40 is a graph of the response of a sensor to a voltage pulse;

FIG. 41 is a graph of disturbance measurements;

FIG. 42 is a graph of measured magnetic field as a function of magnet position;

FIG. 43 is a schematic diagram of a control system embodiment of the present disclosure;

FIG. 44 is a schematic diagram of a feedback loop of the present disclosure;

FIG. 45 is a detailed schematic diagram of the control system embodiment of FIG. 43;

FIG. 46 is a flow chart of the operation of the control system of the present disclosure in the sensing mode;

FIG. 47 is a flow chart of the operation of the control system of the present disclosure in calibration mode;

FIG. 48 is a hysteresis diagram of a micropump embodiment of the present disclosure; and

FIG. 49 is a graph of set points and positions in the closed loop control system of the present disclosure;

FIGS. 50 through 52 illustrate check valves that may be used with the micropump of the present disclosure;

FIG. 53 is an exemplary embodiment of a multi-chamber micropump of the present disclosure.

Detailed Description

Referring now to fig. 1 and 2, a micropump of the present disclosure includes a pump assembly 10 having a first pump body 12 and a second pump body 24 and a flexible diaphragm 36 disposed between the two pump bodies. The first pump body 12 defines a first body flow path and includes a first chamber 14 having a first chamber wall 16 and a first side wall 18. The first pump body 12 also includes a first inlet port 20 and a first outlet port 22 in fluid communication with the first chamber 14. Similarly, the second pump body 24 defines a second body flow path and includes a second chamber 26 having a second chamber wall 28 and a second side wall 30. The second pump body 24 also includes a second inlet 32 and a second outlet 34 in fluid communication with the second chamber 26.

The micropump of the present disclosure also includes a driver magnetically coupled to the flexible membrane 36. In the embodiment illustrated in fig. 1 and 2, the driver includes a first magnetic coil 38 and a second magnetic coil 40. The magnetic coils 38, 40 are configured to exert a magnetic force on the flexible diaphragm through electromagnetic coupling with the magnets 42, 44.

The micropump of the present disclosure is used to make a fluid flow in one direction. This one-way flow is achieved with or without a check valve. Typical operating flow rates for micropumps are in the range of about a few microliters to a few milliliters per minute (below 10 μ l/min for non-mechanical micropumps and up to a few milliliters for mechanical micropumps). Thus, a wide range of biomedical applications is found in applications such as fluid fine management and accurate control systems for implantable drug delivery, chemical and biological detection, and blood transfusion in cardiology systems.

However, there are problems associated with check valves, such as high pressure losses, susceptibility to solid particles, and wear and fatigue of movable valves. Thus, to eliminate the need for check valves, a nozzle/diffuser configuration may be used in place of a check valve and to regulate flow. Thus, a micropump that uses the difference in flow resistance through the nozzle/diffuser element to direct the flow in a preferred direction is referred to herein as a "valveless micropump".

In exemplary embodiments of the present disclosure, one-way rectified fluid flow may be achieved without check valves by using nozzle/diffuser passages at the inlets 20, 32 and outlets 22, 34. The features of the valveless embodiment will be explained with reference to fig. 3, which shows an embodiment of the second pump body 24. Obviously, the first pump body has the same features and is omitted for clarity. In this embodiment, the inlet 32 and the outlet 34 include an inlet diffuser 46 and an outlet diffuser 48, respectively, that are in fluid communication with the second chamber 26.

With particular reference to the inlet diffuser 46, the diffuser element includes a pair of walls 50, 52 connecting the second inlet 32 with the second chamber 26. The walls 50, 52 are disposed at an angle θ and have a length L. The walls 50, 52 define an inlet throat 54 and an outlet end 56, wherein the inlet throat 54 has a width W₁The outlet end 56 has a second width W₂And W is₂Greater than W₁. In the embodiment shown in FIG. 3, the inlet and outlet diffusers 46, 48 have the same depth as the second chamber 26, which has been found to simplify manufacture, but other configurations are acceptable, including the frusto-conical configuration shown in FIG. 4 and the frusto-pyramidal configuration shown in FIG. 5And (4) a shape structure.

The frusto-conical diffuser of FIG. 4 includes a diameter D₁Inlet throat 54 and diameter D₂Outlet end 56 of (2), wherein D₂Greater than D₁. The frusto-conical diffuser also includes a wall 58 disposed at an angle of 2 θ. Similarly, the truncated pyramid diffuser of FIG. 5 includes a cross-sectional area A₁Inlet throat 54 and cross-sectional area a₂Outlet end 56 of (a), wherein A₂Greater than A₁. The frusto-pyramidal diffuser further comprises wall segments 60 disposed at an angle of 2 theta.

For simplicity, FIG. 6 shows a schematic of a micro-pump having a single chamber 62, a single magnet 64, and a single electromagnetic coil 66 powered by a power source 68. Fluid contained in the fluid reservoir 70 flows through the inlet pipe 72 to the inlet diffuser 46, into the chamber 62 where it is pumped through the outlet diffuser 48 through the outlet pipe 74 for its intended use.

Referring to fig. 7, an embodiment of the valveless micropump of the present disclosure may include a combined parallel flow path, wherein the dual chamber micropump 10 is configured to have the first chamber 14 and the second chamber 26 in fluid communication with the common inlet 76 and the common outlet 78. Of course, it is apparent that the embodiment of the micro-pump 7 shown in fig. 7 may be configured with independent parallel flow paths. The independent parallel flow paths can allow simultaneous flow of two different fluids.

Referring now to fig. 8 and 9, in another embodiment of the present disclosure, a micropump 10 as previously described is included as part of the device 200 and enclosed in a housing 202, 204. The housings 202, 204 are configured to include a controller (not shown). The controller is connected to a control panel 206 to allow a user to input operating parameters, such as flow rate. The control panel 206 includes a display 208 and one or more input buttons 210. The housing 204 is configured to receive a bottle 212 that serves as a fluid reservoir for the micro-pump 10. In an exemplary embodiment, the bottle 212 may contain insulin, or any other medication, biological substance, or compound. The housing 204 is configured such that the bottle 212 is in fluid communication with the micro-pump 10 when inserted into the housing. The housing 204 is also configured to receive a battery 214, which serves as a power source for the actuator and controller. In the embodiment shown in fig. 9, the battery is described as a standard 9V battery. However, depending on the application, other types of batteries may be acceptable, for example a 3V coin (watch) battery may be used in some applications where overall size is a consideration.

Referring now to fig. 10, the micro-pump 10 of fig. 1 may be configured as a pump cartridge (pump cartridge)80 that is insertable into a driver 90. The pump cylinder 80 comprises a first pump body 12, a second pump body 24, and a flexible diaphragm disposed between the two pump bodies. The pump barrel 80 optionally includes a check valve manifold 82. Alternatively, the pump cartridge 80 may have a valveless design as disclosed herein. The inlet and outlet pipes 72, 74 are then connected to the check valve manifold or, in the case of valveless micro-pumps, directly to the inlet 20, 33 and outlet 22, 34 ports of the first and second pump bodies.

The driver 90 includes a first support 92 and a second support 94, the second support 94 being disposed separately from and spaced apart from the first support 92. The first and second supports 92, 94 each include a recess 96, 98, respectively, configured to receive a solenoid or actuator coil (not shown). The first and second supports 92, 94 define a receptacle 100 configured to receive the pump barrel 80.

A number of proposed actuating mechanisms for micropumps have been reported, including mainly piezoelectric, electrostatic, electromagnetic and thermo-pneumatic actuating mechanisms and shape memory alloys, among others. Most micropumps use piezoelectric or electrostatic actuation, which operates at relatively high frequencies and require hundreds to thousands of high voltages for extremely small displacements. With respect to electromagnetic actuation, it has advantages over other actuation methods when large displacements, fast response times and relatively low power consumption are highly desirable. Magnetic actuation of a diaphragm with integrated magnets can produce hundreds of μ N and large diaphragm deflections. These desirable characteristics are highly desirable for many medical applications. Accordingly, the effect of the fluid-diaphragm coupling on the response frequency of the solenoid-driven valveless micropump is discussed in detail in the following section.

An actuation force is applied through the oscillating diaphragm to drive a working medium in the pump. Therefore, the reliability and performance of the micro-pump is dependent on the dynamic characteristics of the composite diaphragm.

For an oscillating diaphragm, material properties such as density, young's modulus and poisson's ratio will significantly affect the natural frequency of the diaphragm. For example, in MEMS devices, most membranes are monolithic composite layers that include some sensing or actuating membrane layers. In this particular example, the properties of the material layers are very different from each other. Therefore, the equivalent density (equivalent density) of the composite layer must be correctly derived.

For magnetically actuated diaphragm micropumps, there are two approaches for creating a functional diaphragm. One is that the soft magnetic material is plated or bonded on top of the membrane with permanent magnets that are manually assembled into the PDMS membrane. An external magnetic field is then applied through a permanent magnet or integral planar micro-coil in the substrate to control the movement of the diaphragm. The composite diaphragm is here manufactured to have magnetic properties, since the size and arrangement of the block magnets embedded in the diaphragm may affect the distribution of electromagnetic forces as well as the diaphragm stiffness.

Silicon, silicon nitride and thin metal plates are suitable as membrane materials for micro-pumps. For example, thin silicon diaphragms in the range of a few microns can be realized using micromachining techniques. However, the young's modulus of silicon is 190Gpa, which limits its use in reciprocating pumps. The pump diaphragm is of a flexible material such as parylene, polyimide, SU-8 and PDMS. These diaphragms require less actuation pressure and have greater deflection and greater stroke volume. In an exemplary embodiment of the present disclosure, PDMS (Silgard184, Dow Corning Corp) is used for both the micro-pump body and the actuator membrane.

PDMS (Sylgard 184silicon Elastomer, Dow Coming Corporation) was chosen as the diaphragm material in this exemplary embodiment due to its lower modulus and good compatibility with silicon and glass substrates. Hard barium ferrite powder (UMBS-IB, united industry co., Ltd, China) was mixed into PDMS (in a 1: 1 weight ratio) to form an actuation diaphragm. The composite diaphragm has homogenous and isotropic material properties and can be deflected bi-directionally in an external magnetic field. The material properties of the composite materials used in the present disclosure are shown in table 1.

Table 1: material Properties of the separator composite

Parameter(s)	Pure PDMS	PDMS composite material	Fe powder
				Young's modulus	1.8e3	2.56e6	2.11e8
Density (Kg/m)3)	1026.9	2053.8	7850

Poisson ratio

0.5

0.5

0.33

The main challenge of manufacturing is to produce thin composite membranes. Thin composite membranes with bulk magnets are prone to cracking when released from the mold during the manufacturing process, but thick membranes suffer from limited deflection under magnetic force. In an exemplary embodiment, a 0.15mm thick layer of PDMS was spin-coated on a silicon wafer and cured at 75 ℃ for two hours. A magnet is disposed in the middle of the first PDMS layer. Then, liquid PDMS was poured around the magnet to form a layer of 0.5mm thickness. A glass slide was used to remove excess PDMS. The septum was placed on a hot plate at 100 ℃ for 30 minutes. Finally, a third 0.15mm PDMS layer was covered on top and cured at 75 ℃ for two hours.

The polarity of the composite diaphragm depends on the polarity of the bulk magnet. Thereby, the magnetic force on the membrane is reversed when the magnetic field is switched. The amplitude and frequency of the oscillating diaphragm is controlled by an AC square wave input current applied to the solenoid actuator. The electromagnetic force is measured directly for diaphragm analysis. The total static electromagnetic force at different currents on the composite diaphragm was measured and is listed in table 2. This shows that the attractive force is greater than the repulsive force because of the reduction of the air gap caused by the attractive force. Thereby, the composite diaphragm will move continuously until an equilibrium is reached between the magnetic force and the elastic force of the diaphragm. The maximum attractive and repulsive forces on the membrane were 23.7mN and 21.7mN, respectively, at a current of 0.2A, which was used to estimate the maximum deflection and stress distribution of the composite membrane by FEA.

Table 2: electromagnetic force on diaphragm

Current (A)	Repulsion force (N)	Suction (N)
			0.10	0.0099	-0.0117
0.12	0.0123	-0.013
			0.14	0.0142	-0.0151
0.16	0.0169	-0.0179
			0.18	0.0189	-0.0204
0.20	0.0217	-0.0237

Resonant frequency

Commercial software ANSYS 10.0 was used to model the composite membrane. Two types of 3D primitive types are mainly used: a body 45 and a housing 64. The entity 45 was used as a base for an embedded bulk NdFeB magnet (thickness: 0.5mm) and a PDMS layer (thickness: 0.5mm) around the magnet. The other two PDMS layers of thickness 0.15mm, overlaid on the top and bottom of the composite structure, are gridded with a cell type housing 63, as shown in fig. 11. There are 1917 nodes and 2208 primitives in this model. The material properties of the separator used for the calculation are shown in table 3. Since the end width of the microchannel (0.38mm) is small compared to the diameter of the membrane (7mm), all fixed edge boundary conditions of the composite membrane are set here.

Table 3: material Properties of the separator composite

Material	Density (Kg/mm)3)	Young's modulus (MPa)	Poisson ratio
				PDMS	1.0269E-6	0.75	0.449
NdFeB	6.667E-6	151	0.24

The trapezoidal cross section and maximum deflection 34.34 μm of the composite diaphragm are observed in FIG. 12. This value is less than the depth of the fluid chamber. Thus, the diaphragm does not contact the bottom of the chamber, particularly when fluid is loaded and a fluid resistance acts on the diaphragm. Due to the different material properties of the composite diaphragm, the magnetic force is concentrated on the bulk magnet area. The stress distribution of the diaphragm is shown in fig. 13. The stress concentration areas are mainly distributed at the four corners of the magnet due to the square shape of the block magnet. The maximum stress in the membrane is about 0.1424MPa, which is less than the shear stress (shear stress) of PDMS material. The static analysis of the diaphragm ensures the safety and reliability of the operation of the micro-pump.

Fig. 14 shows first and second model shapes of a flexible diaphragm. It was observed that the diaphragm was bent in one direction in the basic model and the peak occurred in the middle region due to the embedded block magnet. There are two peaks for the second model, one above and one below. The first model is preferred because of the maximum stroke volume required to obtain high pumping flow rates. This analysis also explains why the flow velocity is less than the former under the second vibration model. When fluid is loaded in the chamber, the resonant frequency is lowered due to the increased mass and damping effect on the dynamic characteristics of the diaphragm.

The approximation of the resonant frequency can be simplified by using a mass-spring simulation in which the elasticity of the diaphragm is represented by the spring and the fluid in the chamber is represented by a mass. However, the difference between the calculated and measured frequencies is as large as 36% (calculated and measured frequencies: 734Hz, 540 Hz; 4238Hz, 3350Hz, respectively). It ignores the non-linear behavior in the nozzle/diffuser element that leads to an overestimation of the resonance frequency. The approximation model relates diaphragm vibration to fluid in the pumping chamber based on a set of partial differential equations. Thickness h and mass density ρ_mThe thin circular diaphragm-fluid coupling control formula is:

wherein D ═ Eh³/12(1-v²) Is the bending stiffness;

is the laplacian in polar coordinates; v and E are Poisson's ratio and Young's modulus, respectively.

The deflection of the membrane is relatively small compared to the characteristic length of the membrane. Thus, the theory of small deflections of the thin plate can still be used in diaphragm micropumps. The plate is assumed to be formed of a linear elastic, homogeneous and isotropic material and the effect of shear deformation is ignored. The scheme adopts the following form:

wherein W_mn(r)＝A_mnJ_m(λ_mnr/R)+B_mnY_m(λ_mnr/R)+C_mnI_m(λ_mnr/R)+D_mnK_m(λ_mnr/R)

Where m and n are the number of pitch circles and diameter lines; a. the_mn、B_mn、C_mnAnd D_mnIs the model shape constant and is determined by the boundary conditions. J. the design is a square_m、Y_mIs a Bessel function of a first and a second type, I_m、K_mAre modified bezier functions of a first and a second type. R is the radius of the diaphragm.

On the fluid side, we consider fluid flow as an incompressible, laminar flow. Also, we assume that the fluid loading does not change the model shape, although it will increase the effective mass and damping. Thus, the Navier-Stokes equation and the mass continuum equation are used to describe the fluid flow within each element shown in FIG. 3.

WhereinIs the viscosity of the fluid in the x, y, z directions. The dynamic pressure p represents the coupling of the diaphragm vibration and the fluid flow during the pumping phase.

The volume flow through the inlet and outlet from the inside to the outside can be denoted Q_nAnd Q_d. The pressure loss can be expressed asWhere ξ is the loss coefficient,is the average flow velocity through the throat region of the nozzle/diffuser element. The deflection w of the diaphragm results in a change in the volume of the fluid, expressed as:

V(p，t)＝∫∫w(r，θ，t)drdθ(5)

the rate of volume change is thus given by:

for the particular case where there is no pressure difference, the input pressure is zero and the excitation force is assumed to be sinusoidal. Solving equations (1) - (6), the original expression for the resonance frequency taking into account the fluid action is derived from [11] and can be rewritten in the form:

A₁＝HW₁ A₂＝HW₂(10)

where β is the ratio between the mass of the corresponding effective fluid and the mass of the diaphragm, which relates to the ratio of the density of the fluid and the diaphragm, the area A of the vibrating diaphragm_mAnd diffuser element size variables (chamber height H, diffuser element length L, and throat section width W, as shown in fig. 3). Wherein f is₀Is the fundamental frequency of the thin plate of the clamping edge. Thus, equation (7) means that the resonant frequency of a diaphragm micropump is related to the morphological properties (modal properties) of the diaphragm, the density ratio between the fluid and the diaphragm, and the geometric configuration and dimensions of the micropump.

The combination of acoustic and structural mechanics models in the COMSOL modeling software (provided by COSMOL AB of Stockholm, Sweden) addresses the issue of coupled fluid-elastic structure interactions. In multiple physical couplings, an acoustic analysis provides load (pressure) to a structural analysis, and the structural analysis provides acceleration to the acoustic analysis. Here, the pressure is related to the density by the speed of sound in the fluid. The diaphragm is assumed to be clamped at the outer edge where the displacement and velocity are zero. When the diaphragm is flexed under an electromagnetic field, the acoustic pressure from the fluid is used as the nominal load. For the fluid part, we assume that the base of the micro-pump is the preferred rigid wall, whereby the nominal acceleration disappears at the wall of the diffuser/nozzle element and at the fluid chamber wall. A non-slip condition is set at the fluid-wall interface and no pressure boundaries are set at the inlet and outlet. All boundary conditions are set as described above.

Equations (7) - (11) teach that many influencing parameters will be influencing factors for the change in the resonant frequency. Thus, dimensionless variables must be established using the platinum-Han's II theorem to identify these factors.

f₁＝f(L/W₁，θ，H/W₁，H/h) (12)

Wherein L/W₁Defined as diffuser aspect ratio; theta isDiffuser opening angle; H/W₁Is the aspect ratio of the diffuser; H/H is the thickness ratio. Thus, to directly illustrate the relationship between the resonant frequency and the geometric influencing parameter of a diaphragm micropump with fluid-diaphragm coupling, after W and h are specified, analytical and numerical solutions are plotted in fig. 15 to 18. Water was used in this example.

FIGS. 15 to 18 show the resonance frequency and diffuser aspect ratio (L/W)₁) The inverse proportional relation between the resonant frequency and the opening angle (2 theta), the height-width ratio (H/W)₁) And the thickness ratio, are observed in both the finite element method and the analytical method. Moreover, the FEA method is well consistent in magnitude with analytical predictions. The analytic solution is within 20% of the FEA solution. For comparison, when air is loaded for testing, the difference can be reduced to as low as 10%. Based on the foregoing analysis, the dimensions of the particular micro-pump model were selected in table 4 and the properties of the room temperature working fluid are listed in table 5.

Table 4: size of the micropump embodiment

Parameter(s)	Numerical value
		PDMS composite diaphragm thickness h (mum)	65
Radius of a (mum) of a circular fluid chamber	3000
		Fluid chamber depth H (μm)	650
Diffuser/nozzle throat width W1(μm)	160
		Diffuser/nozzle tip width W1(μm)	440
Diffuser/nozzle opening angle (2 theta)	10
		Diffuser/nozzle depth HI (mum)	650
Diffuser/nozzle length L (μm)	1600

Table 5: parameters for room temperature working fluid calculation

Parameter(s)	Air (a)	Water (W)
			Speed of sound (M/s)	343	1500
Density (Kg/m)3)	1.2	1000
			Viscosity (N.s/m)2)	1.8e-5	0.001

The first two resonant frequencies of the actuator diaphragm when no fluid is loaded are about 138.106Hz and 287.222 Hz. When air was loaded, the resonant frequency was slightly lowered to 104.762Hz and 284.198 Hz. When water was used for the test, the frequencies were 5.531Hz and 65.269Hz, respectively. This comparison shows that the increase in density adds mass to the system, resulting in a decrease in the resonant frequency, and that the higher the fluid density, the more pronounced the damping effect. It can be observed that a circular diaphragm bends in one direction at a first resonance frequency and has a peak in the middle of the diaphragm, which is preferred in micro-pump actuation. This is consistent with the assumption that fluid loading does not change the model shape (modal shape).

Another observation to demonstrate that fluid damping occurs during pumping action is diaphragm displacement. A comparison of the transient analysis of the micropump over a period of time with or without fluid damping to actuate the diaphragm is shown in fig. 19. At 0.05s and 0.15s, the maximum displacement of the diaphragm occurs in the opposite direction and the microdevice is in the pump and supply modes, as shown in fig. 20 and 21, respectively. The amplitude of deflection of the actuator diaphragm at an excitation current of 0.4A was 87.691 μm, which is less than 104.5 μm when no fluid was coupled in the chamber. Thus, the deflection amplitude of the diaphragm was reduced by 16.09%. This again means that the fluid damping effect occurs during the pumping action.

Rigid fluid chambers with microchannels, and flexible actuating diaphragms for such devices, are desirable for better performance and reliability. The soft and flexible polymer chamber will cause vibrations throughout the entire micropump. Increasing the percent of curing agent in the mixture increases the stiffness of the PDMS. Thus, the curing agent in the PDMS mixture is increased to a ratio of about 5: 1 PDMS to curing agent to provide a rigid structure to the fluidic chamber. This ratio for the membrane was 10: 1 for PDMS to curing agent. Then, liquid PDMS is poured into the SU-8 mold and cured to obtain the desired microstructure. Finally, the two layers are carefully aligned and pressed together.

With respect to bonding techniques, normal operating conditions, such as room temperature and normal pressure, are preferred for low cost manufacturing. As multi-layer PDMS microdevices have attracted interest over the last few years, a number of different PDMS bonding techniques have been reported and compared with respect to their bonding strength. Fast, but expensive oxygen ion bonding techniques are still widely used methods for bonding PDMS layers, while uncured PDMS adhesives provide an effective and simplified alternative to oxygen ion bonding. Both of these approaches are acceptable for assembling the actuator diaphragm and fluid chamber base of embodiments of the present disclosure. A very thin film of uncured PDMS was applied to the surface of the molding fluid chamber PDMS substrate on a 100 ℃ hot plate for 20 minutes. Alternatively, oxygen ion treatment (Spacemaker)Microwave oven, 10% oxygen for 10 seconds) also provides a very robust bonding method between the two PDMS layers to seal the fluidic chambers and the inlet/outlet microchannels. The weight of the micropump was measured to be about 1.47 g.

Another challenge faced after the micropump is assembled is the interconnection between the microdevice and standard fluid equipment (e.g., large syringes and tubing). Microfluidic technology involves dimensions on the order of millimeters or less and thus there are no readily available micro fluidic connections to accommodate tubes of different dimensions. The PDMS connector 300 (which includes a hole through the center and double-sided glue for connection to the microfluidic device) and the plastic assembly 302 were fabricated by CNC machining, as shown in fig. 22. Fitting 302 includes a small end 304 that is connected to a tube 306 and another tapered end 308 that is pressed into the soft polymer device.

The average volumetric flux over a period of time is one of the most important characteristics of a micropump. Inertial effects and loss of capacity in the nozzle/diffuser element and fluid chamber, as well as loss in the actuating diaphragm, are considered herein. The frequency dependent flow rate is derived based on fluid-diaphragm coupling control equations (1) - (4) and fluid volume equations (5) - (6).

Wherein:

C＝C₁+C₂+α²C₄，ω₀＝2πf₀(14)

item C₁Represents the inertial effect of the diaphragm; c₂The inertial force of the fluid in the pump is considered; c₃Reflecting viscous losses in the nozzle/diffuser element; c₄Representing the inertial effects of the fluid within the nozzle/diffuser element; c relates all inertial factors of the fluid and the membrane. F is a dimensionless actuation force and there is no pressure difference between the inlet and the outlet. Item C if a parallel dual fluid chamber configuration is used, as shown in FIG. 10₂、C₃And C₄And is doubled accordingly.

The theory obtained above indicates that the pressure loss coefficient (ξ)_n/ζ_d) The ratio between should be as high as possible to maximize the pumping stroke efficiency. Thus, for each pumping cycle, the nozzle/diffuser pump may produce a net flow from the nozzle to the diffuser. From these equations, we observe that the flow rate is affected by three factors: ratio of excitation frequency to diaphragm fundamental frequency (ω/ω)₀) Density ratio (R)_ρ) And a geometric ratio that determines the loss coefficients (α and β). However, it should be noted that the pressure loss coefficient (ξ)_n1.01 and xi_d0.449) is numerically obtained at low Reynolds number (Reynolds number) using Finite Element Analysis (FEA), since the pressure drop depends mainly on dimensionless variables according to the platinum han II theorem. And thirdly:

ξ＝f(L/W₁，θ，H/W₂，1/Re)(15)

at the same time, to some extent, the surface roughness in the nozzle/diffuser element also affects the pressure drop. However, if the micro-pump has been manufactured and the flow rate can be measured directly, it is not necessary to accurately measure the geometry of the micro-nozzle/diffuser and calculate the loss factor. FEA is therefore an efficient method of calculating the loss factor and predicting the pumping flow rate in the conceptual design stage. Based on the analysis in section 3 and the actual requirements of the excitation frequency, a valveless micropump with the desired dimensions shown in table 4 was used to study the frequency dependent performance.

The graph shown in fig. 23 illustrates that the pumping flow rate of the micro-pump is a function of the excitation frequency, which varies from 0Hz to 20 Hz. In the low frequency range, the flow rate increases almost linearly with the excitation frequency and then reaches a maximum flow rate at the resonant frequency of the actuated diaphragm with fluid damping. After the flow rate peak, the pumping rate decreases sharply at high frequencies. Referring to fig. 24, the maximum pumping rate increases linearly with the actuation voltage amplitude. The maximum flow rates at 0.4A and 2A were 19.61. mu.l/min and 43.86. mu.l/min, respectively. An increase in the amplitude of the voltage leads to an increase in the deformation of the diaphragm. To meet high flow rate requirements, for example, in a drug delivery system, maintaining the same input power, a parallel dual chamber configuration is shown in fig. 7. Interestingly, however, although the fluid capacity was doubled for the latter (which operated in anti-phase), the maximum flow rate was about 27.73 μ l/min at 0.4A input current with an excitation frequency of 3 Hz. This is less than twice 19.61 μ l/min for the former operating at the same current amplitude but with an excitation frequency of 4.36Hz (as shown in figure 25). This result is also reasonable since the fluid in both chambers will act.

In an exemplary embodiment of the micropump of the present disclosure, a low cost simple solenoid is developed for magnetic actuation to replace the integrated micro-coil to avoid complex and rigorous manufacturing processes. Despite the limitations of the structure of the external magnetic actuator on its application, the electromagnetic drive has more than it has in cases where high force, fast response and low power consumption are highly required and size is a secondary considerationHe motivates the advantages of the method. Simple in design and easy to manufacture, the electromagnet comprises a magnetic induction coil wound around a soft iron cylindrical rod (5mm diameter X10mm), and a NdFeB magnet (size: 3 x3x0.5mm) with integrated small blocks³And weight 0.03g, from neotex, berlin, germany). The electromagnetic drive can directly generate a controllable magnetic field from the actuation current, although typically weak over very short distances. Thus, when the magnetic field is reversed, alternating attractive and repulsive forces are generated on the composite diaphragm, which generates a periodic flexing of the diaphragm.

The resistance and inductance of the copper coil (28AWG, 460 turns) were about 4.40 Ω and 3.49mH at 100 Hz. Currents in the solenoid above 0.5A generate heat very quickly. Therefore, the actual current of actuation should be controlled below 0.5A. Two types of actuation currents are typically used: sine wave and square wave currents. With the same peak-to-peak value, a square wave current can hold a large deflection and carry more energy, which can be converted into a magnetic force. A DC power supply (which can provide a maximum-30/+ 30V voltage (BK Precision 1672)) and a square wave generator circuit can be used to generate the square wave current, as shown in fig. 25A. After the solenoid actuator is loaded due to the induction of the coil, a small change in the square wave signal occurs. The DC power supply, which can be replaced by a battery, has the possibility of being used for portable applications of the micro-pump. The frequency of the signal is controlled by the resistance of the potentiometer in the fine adjustment circuit.

The vertical electromagnetic force Fz exerted by the induction coil on the permanent magnet is given by:

wherein H_zIs the vertical component of the magnetic field generated by the coil, B_rIs the remanence of the magnet, S_m，、h_mRespectively the surface area and the thickness of the magnet.Is the gradient of the magnetic field. The formula indicates that the electromagnetic force is proportional to the magnet volume and the change in the vertical magnetic field.

In one exemplary embodiment of the micropump of the present disclosure, the design parameters listed in table 6 are used.

Table 6: structural size of micro pump

Design parameters	Mini pump (mum)
		Thickness h of the diaphragm	800
Radius of fluid chamber a	3500
		Fluid chamber depth H	500
Diffuser/nozzle throat width (W1)	160
		Diffuser/nozzle tip width (W2)	440
Diffuser/nozzle opening angle (θ) degree	10
		Diffuser/nozzle length (L)	1600

Depending on the flow rate requirements in the application, the volume stroke and excitation frequency for a particular diaphragm can be estimated and thus the magnetic and electrical input signals required for actuation can also be obtained in the design phase. However, it is very impractical to estimate these parameters by experimentation. Therefore, a frequency dependent flow rate formula is used to roughly estimate these parameters, since the geometry of the diaphragm, the fluid chamber structure, the microchannels and the fluid properties determine the resonant frequency and thus the flow rate of the micropump.

Q＝2ηΔVf(17)

Where, Δ V is the stroke volume,is defined as the pump stroke efficiency, and η_FIs the diffuser rectification factor and f is the excitation frequency.

Then, in the apparatus, the design parameters were determined as in table 1. Notably, increasing the stroke volume and reducing the dead volume improves the performance of the pump. The total volume of the pump was 0.01924ml, with a diameter of 7mm and a depth of 500 μm.

For the exemplary embodiments of the present disclosure, ethanol is used as the working fluid. The physical properties of the media at 20 ℃ and 1 atm are listed in table 7. In this example, the device includes a fluid reservoir (syringe), a micro-pump, an actuator circuit board, and an electromagnetic actuator, and an optical microscope with a CCD camera. A microscope was used to observe the ethanol within the fluid chamber and the bubbles generated during the pumping process.

Table 7: fluid properties at 20 ℃ and 1 atm

The inlet and outlet pipes being of commercial useA tube. The inlet tube is connected to the reservoir. The fluid reservoir automatically fills the fluid chamber as the fluid moves forward. The self-filling capacity and the bubble tolerance can be determined according to the compression ratio (the ratio between the volumetric stroke av of the pump and the total dead volume V). Since the total volume V of the micro-pump is constant when the micro-pump has been manufactured, the volume stroke determines the compression ratio. In this case, the compression ratio is about 0.068, which is less than the minimum compression ratio of 0.075 for a self-priming and bubble tolerant fluid pump.

During operation of the micropump, the fluid flow rate in the outlet pipe and the weight of the fluid are measured. The friction of the tube against the fluid flow is of particular importance. The pressure drop in the tube must be taken into account. The flow rate of interest for micropumps for biomedical applications is typically less than 1 ml/min and the reynolds number can be estimated to be 8.72. Thus, the flow through the tube is laminar. The change in pressure due to frictional losses in the inlet/outlet pipes can be evaluated by the Hagen-Poiseuille equation:

where Δ p is the flow resistance and μ is the fluid viscosity, Δ L, a is the length and inner radius of the tube, and Q is the flow rate.

The pressure drop of each fluid medium can be neglected (the length of the tube under test is about 5cm, the pressure drop is about 2.1Pa) and does not significantly affect the performance of the micro-pump.

The maximum flow rate is the flow rate when the pump is operating at zero back pressure. For these different current tests, there was no pressure difference between the inlet and the outlet. Micro-pumps operating at resonant frequencies can result in increased displacement, higher flow rates, and higher conversion efficiency, thereby reducing power requirements. Therefore, the effective excitation frequency of the control system becomes very important. Currents of 0.14A, 0.16A and 0.18A were used to test the flow rate over the excitation frequency range. These results are shown in FIGS. 26-28.

When the frequency is below 15Hz, the flow oscillates around a position near the beginning of the outlet pipe connected to the micro-pump and the pump cannot deliver the fluid. This is because of the low fluid resistance in both directions and thus the valveless pump experiences some degree of backflow. Therefore, if the excitation frequency is too low, it is difficult to accumulate sufficient net fluid flow from the inlet to the outlet. Furthermore, when the drive frequency is below 20Hz in a valveless rectification pump, the fluid flow is pulsating and it is difficult to maintain a constant flow rate. This is due to the periodic nature of the square wave signal applied in the magnetic actuation. Also, for these three sets of examples, there are two flow rate peaks. This is because the first two modes of the resonant frequency of the vibrating membrane (the first two modes of resonant frequencies) are achieved.

As shown in fig. 14, the diaphragm is curved in only one direction, with half of the diaphragm being curved upward and the other half being curved downward. The first mode produces a higher volume stroke than the second mode. Thus, the flow rate is typically lower at the second peak than at the first peak.

There were some differences in the flow rate curves in these three sets of tests with different current amplitudes. First, for the first set, the effective operating frequency range for a stable flow rate is from 20Hz to 34Hz, and the two resonant frequencies are 25.01Hz and 30.04Hz, respectively. 20Hz to 47.5Hz and 20Hz to 50Hz are the frequency ranges of the second and third group, respectively. The two resonance frequencies are 25.9 and 36.1Hz for the second group and 26.1 and 37.5Hz for the third group. The values of the latter two groups are very close while the first group deviates slightly. Since the flow rate is small at a current of 0.14A and the friction force in the tube becomes the dominant factor, the measurement error is higher than the other two groups. Second, the flow rate increases with the excitation frequency before the first flow rate peak is reached, followed by a sharp decrease. The second peak, which is slightly lower than the first peak, comes again with an increase in the excitation frequency. Third, the flow rate decreases as the frequency increases. The results show that the flow rate can be controlled within the relevant excitation frequency.

Backpressure generally refers to the pressure in the fluid system created by the obstruction against the free-flowing flow. Thus, in the micropump of the present disclosure, the maximum back pressure (P)_max) Is defined as the counter pressure exerted on the fluid when the flow rate of the pump becomes zero.

Achieving a constant continuous pumping over a long period of time is an important indicator of a reliable micro-pump. The rise in temperature in the fluid chamber is an important feature because small bubbles generated in the chamber can significantly affect the flow rate. Furthermore, in biochemical or life science applications, high temperatures can damage fluids containing living cells or sensitive particles. However, it is not easy to measure the temperature change in the fluid chamber. Instead, the temperature rise of the magnetic actuator is measured to estimate the fluid temperature, since the magnetic actuator is current driven and the temperature of the coil will be rapidly increased. The actual temperature of the fluid may be slightly lower than the temperature of the magnetic actuator. Thus, at 0.18A, 25Hz (resonant frequency) and an air gap of about 1mm between the electromagnet and the vibrating diaphragm, the temperature rise within one hour was measured. As shown in fig. 29, the temperature steadily increased from 21.3 ℃ to 38.1 ℃ along a non-linear curve. This temperature is significantly below the critical temperature for most biological fluid samples. Characteristics of embodiments of the micropump of the present disclosure are listed in table 8.

Table 8: micro pump characteristics

Outer dimension	11x7x2.5mm3
		Actuation	Electromagnetic actuation
Particle tolerance	Is that
		Resonant frequency	25.9Hz
Maximum flow rate	75.13. mu.l/min
		Maximum back pressure	400Pa
Electric current	0.18A
		Voltage of	2.1V
Power consumption	378mW

Thus, an embodiment of a magnetically actuated soft Polymer (PDMS) micropump is given in this example. The fluid flow direction depends on two nozzle/diffuser elements, which have different fluidic resistances in the inlet and outlet of the microdevice. There are a number of advantages associated with this micro-pump embodiment. This simple manufacturing process and planar structure feature allows easy integration into the μ TAS device, thereby allowing miniaturization of the entire microfluidic system. All manufacturing processes can be implemented outside of clean room facilities, which significantly reduces costs. Furthermore, the need for low voltage and power consumption makes the micropump of the present disclosure useful in portable medical devices, which may be powered by small batteries. Based on the given design and manufacturing method, the constant continuous flow rate and low temperature ramp over a long period of time demonstrates the feasibility and good reliability and biocompatibility of the soft PDMS micropump in biomedical applications.

In the micropump of the present disclosure, the control system includes a sensor and a controller. The sensor is a hall effect sensor and is disposed proximate the actuator coil. Referring to fig. 30-32, the flexible diaphragm moves in response to a magnetic field (B) applied by an electromagnetic coil. The position of the magnet and hence the deflection of the diaphragm and the respective volumes of the two chambers change the magnetic field configuration which is sampled by the hall effect sensor. A suitable sensor is, for example, an A1301 Linear Hall Effect sensor manufactured by Allegro Microsystems having a sensitivity of 2.5 mV/Gauss. The position information is provided to the controller and used to determine the position of the magnet (within an accuracy of 0.05 mm).

The controller indicates movement of the magnet based on a user selected flow rate requirement. Multiple modes of operation may be configured, such as a low speed mode for accurate dosing or a high speed mode for high volumetric flow rates.

The ability to measure the real-time position of the magnet is important because it enables closed-loop flow rate control, which prevents collisions between the magnet and the chamber wall (which eliminates collision damage and reduces noise), and enables an efficient controlled resonance mode of operation. Since the micropump of the present disclosure includes two separate parts, a contactless sensing system is necessary. It has been found that determining the position of the magnet, and thus the position of the diaphragm, is easily achieved in a cost-effective manner by measuring the magnetic field generated by the magnet. The main disadvantage of implementing this method is the magnetic noise due to the electromagnetic driver coil, which needs to be suppressed. Thus, the sensor is positioned such that magnetic noise is minimized.

The amplitude of the signal and magnetic perturbations (magnetic perturbations) strongly depend on the position and orientation of the sensor. At the sensor location, the coil magnetic field must be as lowest as possible and the magnet magnetic field must be as highest as possible. There are locations that satisfy these two requirements (see fig. 33, 34, 35). Along the side of each coil, Br_coilWeaker and Br_magnetIt is stronger. This position is called the lateral position. This position provides a better signal to noise ratio for magnetic field induction than would otherwise be possible. In fig. 36, the magnetic fields generated by the coil and the magnet are compared: (i) br at the lateral position and (ii) Bz at the classical standard position (i.e. in the centre of the coil).

This side position proof is for coil B_coilIs very insensitive to the magnetic field of magnet B and still_magnetAnd (4) sensitivity. On fig. 37, it is shown that the signal at this lateral position is 3 times smaller than at the center; however, the perturbation of the coil is about 12 times smaller, so that the signal/noise ratio is better than at this lateral position.

In an exemplary embodiment, the minimum signal-to-noise ratio (B)_magnet)_min/(B_coil)_max45, and a sensitivity ratio Δ B_magnetΔ z 75-100Gauss/mm (depending on location)) 220mV/mm has been determined using the position Z of the magnet (taking into account the maximum distance of 8mm and the maximum current through the coil of 0.3A). Performance and coil perturbations have been measured by the sensor at this lateral position as shown in fig. 38, 39.

This selected position allows to optimize the signal/noise ratio instead of completely suppressing the magnetic disturbances of the coil. To achieve this accurately, a noise suppression system is constructed that is based on the separation of the coil and magnet components of the magnetic field, since pulses are used.

When the step voltage is t ═ t₀Is applied to the coil, the system response is as follows:

B_sensor(t)＝B_sensor(t₀)+ΔB_coil(t-t₀)+Δ_{bmagnet_displacement}(t-t ₀)

wherein Δ B_coil+ΔB_{magnet_dispiacement}Is the total change Δ B in the radial magnetic field measured by the sensor after the step voltage has been applied_sensor。ΔB_coilDue to magnetic perturbations caused by the coil, which is the contribution of the coil to this change (due to the magnetic field generated by the coil); delta B_{magnet_displacement}Is the contribution of the magnet to this change (due to the movement of the magnet and thus the change in distance to the sensor and thus the change in magnetic field measured by the sensor).

These two terms have different intrinsic response times. Delta B_coilProportional to the current I flowing in the coil and thus having an electrical response time τ_I＝L/R。ΔB_{magnet_displacement}Due to the displacement of the magnet. Once the magnetic field is applied, the magnets accelerate until they reach a nearly constant velocity. The time required to reach this constant speed will be the mechanical response time τ_M. In this example, τ_I＜＜τ_M. This means that when a voltage step is applied, Δ B_coilHas reached the final value and Δ B_{magnet_displacement}It is still negligible. Thus, we obtain:

wherein t is₁＞τ_IAnd t is₁＜＜T_M. In this case, t₁0.2ms is an ideal value.

Equation (19) shows that even when the magnet is inside the pump, by measuring the time t, due to the application of the pulse signal₁By measuring the response, the perturbation of the coil can be obtained independently before the magnet has reached its travel.

Based on this principle, the response of the sensor to the voltage pulse (see FIG. 40) is measured and used to calculate the coil Δ B_coil(I) Magnetic perturbation (see fig. 41). The sensor response is compared to its actual valueThe actual value is measured directly from the coil itself by removing the magnet from the pump (see figure K). Note that the measurement is noisy because of the inherent sensor noise level and the limited CAN conversion accuracy at that accuracy level (1Gauss — 1/2000 for the sensor measurement range).

The following protocol results in repeatable accurate values (see fig. 41): applying 5 pulses of +10V for a duration of 1ms, once every 10 ms; applying 5 pulses of-10V for a duration of 1ms, once every 10 ms; the current and magnetic field are measured. The average coil perturbation factor for a pulse is calculated according to the equation:

in A_coilAfter being calculated, A_coilIs used in real time during the measurement, wherein:

B＝B_measured-A_coil*I

the proposed perturbation measurement method is both accurate and fast and can be used in situ with existing magnets in micropumps. This is the first part of the automatic calibration process of the sensing system. The second section is described below.

Once the coil perturbations are suppressed, the measured signal corresponds to the magnetic field B generated by the magnet at the sensor level_rmagnet. The measured signal is not a linear function of the magnet position. Thus, a position determination algorithm is used which transforms the sensor signal into the magnet position.

The magnetization M of the magnet generates a magnetic field at a point r' outside the magnet:

where r is the volume V of the magnet_magentThe coordinates of the elementary volume dV in (20) are used in the volume integration in (20).

In this example, the magnets have an axisymmetric geometry such that for each magnet:

where the reference (0, 0, 0) is the center of the pump chamber. B is_remIs the remanent magnetic field within the magnet. (r)_s，z_s) Is the coordinates of the center of the sensor; r is the radius of the magnet, h is the thickness of the magnet, and Z_bottomIs the coordinate of the bottom of the magnet.

Once the effect of the coil is suppressed, the magnetic field measured by the sensor is a superposition of the magnetic fields of the two magnets.

Equation (21) cannot be used by itself to determine the position z of the center of the magnet_m(and thus the position of the diaphragm) the inverse of (21) cannot be obtained because its integration does not give any analytical result. A look-up table and corresponding counter-look-up table must be established (21).

The actual system is not ideal and the parameters of the formula are known with only limited accuracy. Thus, as B_measuredZ of a function of_mWill differ from the actual value. However, the difference between actual and simulated values can be strongly reduced by using the following equation:

(22)

equation (22) enables the position of the magnet to be accurately determined using corrected analog values that closely approximate the actual values, assuming we know the actual Br in both cases_magnet: for z_BmaxAnd z_Bmin，z_BmaxAnd z_BminAre easily obtained because they are the position of the magnet when it hits the lower wall (z)_Bmax) And position at the time of upper wall (z)_Bmin)。

Following this principle, the position of the magnet can be obtained by the following method:

(a) depending on the size of the pump, the position of the sensor and the size and material of the magnet, the simulation is done off-line by equation (21);

(b) a look-up table is generated and recorded in the microcontroller memory during the burn program;

(c) each time the pump is turned back on or a new pump part is inserted, the maximum and minimum positions are searched and the corresponding magnetic fields are measured;

(d) equation (22) is used to modify the look-up table and then build the inverse look-up table, which is given as B_measureZ of function_m(ii) a And

(e) real-time slave B using linear regression on inverse look-up tables_measuredObtaining z_m；

Wherein steps c and d are performed automatically and do not exceed 1 second.

The magnetic field established by the magnet is measured by the sensor as a function of the position of the magnet (see fig. 42). In fig. 42, the magnetic field is plotted as a function of: (i) a simulated magnetic field; (ii) a corrected analog value according to equation (22); and (iii) real-time linear regression on a look-up table created using the method described above.

The maximum error between the actual value and the calculated inverse look-up table value has been measured: error_max0.03mm is 0.75% of the total range of the magnets within the pump, i.e. with an accuracy of 0.75% of the chamber volume.

Control system 300 is included in an exemplary embodiment of the present disclosure. Referring to fig. 43, the control system 300 includes a Central Processing Unit (CPU)302 configured to receive user input 304. A display 306, such as a Liquid Crystal Display (LCD), is provided to allow a user to see various parameter values, such as flow rate, volume, power, battery charge, etc. The CPU 302 provides actuation signals that pass through a digital-to-analog (D/a) converter and signal processing unit 308. In addition to the Hall effect position sensor 310 near the magnetic coil, as previously described, embodiments of the micropump of the present disclosure may further include a flow sensor 312 on the outlet of the micropump, and a volume sensor 314 on the fluid reservoir. These values are passed through an analog to digital (a/D)316 controller before being provided to the CPU 302. The values obtained from these sensors are compared to provide additional feedback to the controller to optimize the flow rate and/or to provide a warning or alert condition when these parameters exceed predetermined values, as shown in fig. 44.

In an exemplary embodiment of the present disclosure, two different PID controllers are used: (i) the first PID comprises a set point for the current I and the output voltage U; (ii) the second PID includes a set point, for the position of magnet x, and an output of current I. The sensing system takes physical input with the signal from current sensor 31 and converts it to the magnet position after noise suppression (as described previously).

An exemplary embodiment of a control system architecture is shown in FIG. 45. The controller 300 may receive inputs from the sensing module 320 and the calibration manager 330 for operating in the sensing mode and the calibration mode. Also, the signal generated by the sensing module 320 may pass through a filter module 340 before being processed by the controller 300.

When operating in the calibration mode, the calibration trigger 356 is either required to be provided through the user interface. Calibration manager 330 controls the position of the magnet on the flexible membrane by sending a calibration signal to PWM converter 352. The position of the magnet is recorded and the controller parameters are modified. Operation is then returned to the controller with updated parameters and calibration. A simplified flow chart of operation in calibration mode is shown in fig. 47.

When operating in the sensing mode, the user interface 350 provides the required flow rate demand, which is then translated by the controller 300 into displacement instructions for the magnet disposed on the flexible diaphragm. These instructions are then used as set points for the controller 300. The controller compares the set point of the position of the magnet with the actual position in real time. Based on the comparison, the controller 300 sends a voltage signal to a Pulse Width Modulation (PWM) converter 352. The PWM converter 352 then converts the voltage signal from the controller into a PWM signal, which is then supplied to an H-bridge circuit 354, which circuit 354 controls the current flowing to the actuator coil. A simplified flow chart of operation in the sense mode is shown in fig. 46.

Since the exemplary embodiments of the micropump of the present disclosure use hall effect sensors to detect the position of the flexible membrane through the magnetic field strength generated by the magnets disposed on the membrane, the signal needs to be converted to a position value. Exemplary embodiments use a look-up table based electromagnetic field module to provide magnet position as a function of magnetic field strength. The electromagnetic field module is calculated based on parameters (e.g., component dimensions and materials used, as previously described) prior to programming the controller. Since the result of these calculations is an approximation of the physical embodiment, the calibration enables the system to sense the actual position/magnetic field relationship with an accuracy of about 0.03 mm.

The feedback signal provided by the sensing module 320 contains noise due to the nature of the PWM driving circuit and the magnetic field used. Thus, the filter module 340 is used to suppress noise in the signal. The filter module includes two filters: a fast filter 342 and a slow filter 344. The fast filter 342 is less accurate but is suitable for operations that are less sensitive to noise vibrations, such as integration. The slow filter 344 has a higher precision; however, this increased accuracy also increases the delay time. Taking this delay into account, the slow filter is adapted to measure the speed of the magnet.

The system has a clear mechanical hysteresis; furthermore, the current and thus the magnetic field are neither linear nor univocal functions of voltage (univocal function) due to heating of the coil, as shown in fig. 48. This prevents the use of an open loop signal to control it. Despite this lag, which generally prevents better results than the lag amplitude, the prototype response is fast (time response to high amplitude setpoint changes) and accurate (static error 0; maximum overshoot error 2%), as shown in fig. 49.

In addition to the valveless micropump embodiments of the present disclosure, as described above, alternative embodiments may use a check valve to provide a one-way fluid flow path. Referring to fig. 50-52, an embodiment of a check valve 400 for use with the micropump of the present disclosure includes a valve body 402 that may be formed from a pair of body members 404, 406 that are joined together. The body members 404, 406 define a chamber 408 having a pair of substantially planar surfaces 410, 412 disposed in spaced apart and facing relation to one another that serve as a seat for a septum 414 disposed within the chamber 408. Check valve 400 also includes an inlet port 416, an outlet port 418 in fluid communication with chamber 408. The diaphragm 414 is configured to float within the chamber 408 and, in the exemplary embodiment, is approximately 20% smaller than the chamber 408. The seat 410, positioned near the inlet port 416, is physical. The seat 412 positioned adjacent the outlet port 418 includes a plurality of holes 420 that allow fluid to pass through. In the exemplary embodiment, such a hole 420 is machined into seat 412 in a rose pattern, however, any pattern that allows fluid to pass through valve 400 while allowing diaphragm 414 to be seated against seat 412 is acceptable.

Referring now to fig. 51 and 52, at any point where the fluid pressure at the outlet port 418 is greater than the fluid pressure at the inlet port 416, a reverse fluid flow condition will occur, as shown by arrow R in fig. 51. This reverse flow acts to oscillate the diaphragm 414 towards the inlet seat 410. After the diaphragm 414 is seated against the inlet seat 410, the diaphragm 414 covers the inlet port 416 to prevent further reverse flow.

At any point where the fluid pressure at the inlet port 416 is greater than the fluid pressure at the outlet port 418, a forward fluid flow condition will occur, as shown by arrow F in fig. 52. This forward flow swings the diaphragm 414 toward the outlet seat 412. Because the outlet seat 412 includes apertures 420, fluid is allowed to flow through these apertures 420 around the diaphragm 414, thereby allowing fluid to flow through the outlet port 418 in the forward fluid flow direction F

Referring now to FIG. 53, another embodiment of the present disclosure includes a multi-chamber micropump 510. Multi-lumen micropump 510 may be configured as a plurality of pump cylinders 580A, 580B, 580C configured to be inserted into driver 590. Although three pump cartridges are shown, it is apparent that more or fewer cartridges are within the scope of the present disclosure and may vary depending on the application of the micropump of the present disclosure. Each of the cylinders 580A, 580B, and 580C are identical to the others, and thus the cylinder 580A will be described as an exemplary embodiment.

Pump barrel 580A includes first pump body 512, second pump body 524, and flexible diaphragm 536 disposed between the two pump bodies. The pump barrel optionally includes a check valve manifold 582. Alternatively, the pump barrel 580A may have the valveless design disclosed herein. The pump barrel 580A also includes inlets 520, 532 and outlets 522, 534. The tubes may then be connected to the inlets 520, 532 and outlets 522, 534 for fluid transfer.

The driver 590 includes a plurality of receiver modules 590A, 590B, 590C corresponding to the plurality of pump cylinders 580A, 580B, 580C. As an exemplary embodiment, the receiver module 590A includes a first support 592 and a second support 594. The first and second supports 592, 594 each include a recess 598 configured to receive a solenoid or an actuation coil, respectively. Each receiver module 590A, 590B, 590C defines a receptacle 500 configured to receive a pump barrel 580A, 580B, 580C. The receiver modules may be configured in a stacked configuration, as shown in fig. 53, or may be configured in other configurations, such as back-to-back, side-to-side, or combinations thereof, depending on the desired application.

The foregoing is considered as illustrative only of the principles of the invention as claimed. Further, since numerous modifications and changes will readily occur to those skilled in the art, it is not desired to limit the invention claimed to the exact construction and operation shown and described, and accordingly, all suitable modifications and equivalents may be resorted to, falling within the scope of the invention as claimed.

Claims (15)

1. A micropump, the micropump comprising:

a pump assembly, the pump assembly comprising:

a first pump body defining a first fluid flow path, the first pump body comprising:

a first chamber comprising a first chamber wall and a first sidewall,

a first inlet and a first outlet, wherein the first inlet and the first outlet are in fluid communication with the first chamber,

a second pump body defining a second fluid flow path, the second pump body comprising:

a second chamber comprising a second chamber wall and a second side wall,

a second inlet and a second outlet, wherein the second inlet and the second outlet are in fluid communication with the second chamber,

a flexible diaphragm disposed between the first chamber and the second chamber; and

an actuator assembly positioned external to the pump assembly and configured to cooperate with the pump assembly, the actuator assembly comprising:

a driver magnetically coupled to the diaphragm, an

A sensor configured to detect a position of the diaphragm,

wherein the driver applies a magnetic force to the diaphragm causing the diaphragm to deflect, and wherein the deflection of the diaphragm causes a change in pressure within the first and second chambers, thereby causing fluid flow; and

at least one valve in fluid communication with each of the first and second chambers, wherein the at least one valve is configured to direct fluid flow in a predetermined direction.

2. The micropump of claim 1 wherein the at least one valve comprises at least one of a first inlet check valve adjacent the first inlet and a first outlet check valve adjacent the first outlet.

3. The micropump of claim 2 wherein said at least one of the first inlet check valve and the first outlet check valve includes a valve diaphragm including a plurality of intersecting slits radiating outwardly from a common point.

4. The micropump of claim 2 wherein said at least one of a first inlet check valve and a first outlet check valve is located within the first sidewall in the first fluid flow path.

5. The micropump of claim 4 wherein said at least one of the first inlet check valve and the first outlet check valve is integrally formed within the first sidewall.

6. The micropump of claim 2 further comprising at least one of a second inlet check valve adjacent the second inlet and a second outlet check valve adjacent the second outlet.

7. The micropump of claim 6 wherein said at least one of a second inlet check valve and a second outlet check valve is located within a second sidewall in the fluid flow path.

8. The micropump of claim 7 wherein said at least one of a second inlet check valve and a second outlet check valve is integrally formed within the second sidewall.

9. The micropump of claim 1 further comprising a first magnet disposed on said diaphragm.

10. The micropump of claim 9 further comprising a second magnet disposed on the membrane, wherein the first magnet is positioned adjacent the first chamber and the second magnet is positioned adjacent the second chamber.

11. The micropump of claim 9 comprising a plurality of magnets disposed on the membrane adjacent to either of the first chamber and the second chamber.

12. The micropump of claim 9 wherein the first magnet is a neodymium-iron-boron rare earth magnet.

13. The micropump of claim 10 wherein the second magnet is a neodymium-iron-boron rare earth magnet.

14. The micropump of claim 1 wherein the flexible membrane is constructed of a soft polymer material mixed with a magnetic material.

15. The micropump of claim 14 wherein the soft polymeric material is polydimethylsiloxane.