US20100239436A1

US20100239436A1 - A thermal pump

Info

Publication number: US20100239436A1
Application number: US11/383,663
Authority: US
Inventors: Ulrich Bonne; Robert Higashi; Tom Rezachek
Original assignee: Honeywell International Inc
Current assignee: Honeywell International Inc
Priority date: 2005-05-17
Filing date: 2006-05-16
Publication date: 2010-09-23
Also published as: DE602006021005D1; US20090184724A1; US20060262303A1; JP2008541134A; EP1882167B1; WO2007050123A2; US7518380B2; US20070113642A1; WO2007050123A9; WO2007050123A3; US7502109B2; EP1882167A2; US7965089B2

Abstract

A thermal pump for moving a sample fluid to and through an analyzer. The pump may have a lack of moving mechanical parts when pumping except for check valves. The thermal pump may have in lieu of each mechanical check valve a thermal or fluid mechanism that effectively operates as a valve without mechanical parts. The present thermal pump may be fabricated with MEMS technology. The pump may be integrated into a concentrator and/or separator of a fluid analyzer chip.

Description

This application claims the benefit of U.S. Provisional Application No. 60/681,776, filed May 17, 2005. This application claims the benefit of U.S. Provisional Application No. 60/743,486, filed Mar. 15, 2006.
The U.S. Government may have some rights in the present invention.

BACKGROUND

The present invention pertains to pumps and particularly to pumps for micro fluid analyzers. More particularly, the invention pertains to thermal pumps in the analyzers.
U.S. patent application Ser. No. ______, filed May 16, 2006, Attorney Docket No. H0008131 (1100.1371101), entitled “An Optical Micro-Spectrometer,” by U. Bonne et al., is hereby incorporated by reference. U.S. patent application Ser. No. ______, filed May 16, 2006, Attorney Docket No. H0009333 (1100.1410101), entitled “Chemical Impedance Detectors for Fluid Analyzers,” by U. Bonne et al., is hereby incorporated by reference. U.S. patent application Ser. No. ______, filed May 16, 2006, Attorney Docket No. H0010503 (1100.1411101), entitled “Stationary Phase for a Micro Fluid Analyzer,” by N. Iwamoto et al., is hereby incorporated by reference. U.S. patent application Ser. No. ______, filed May 16, 2006, Attorney Docket No. H0012008 (1100.1413101), entitled “A Three-Wafer Channel Structure for a Fluid Analyzer,” by U. Bonne et al., is hereby incorporated by reference. U.S. Provisional Application No. 60/681,776, filed May 17, 2005, is hereby incorporated by reference. U.S. Provisional Application No. 60/743,486, filed Mar. 15, 2006, is hereby incorporated by reference. U.S. patent application Ser. No. 10/909,071, filed Jul. 30, 2004, is hereby incorporated by reference. U.S. Pat. No. 6,393,894, issued May 28, 2002, is hereby incorporated by reference. U.S. Pat. No. 6,837,118, issued Jan. 4, 2005, is hereby incorporated by reference. U.S. Pat. No. 7,000,452, issued Feb. 21, 2006, is hereby incorporated by reference. U.S. Pat. No. 4,944,035, issued Jul. 24, 1990, is hereby incorporated by reference. U.S. Pat. No. 5,876,187, issued Mar. 2, 1999, is hereby incorporated by reference. U.S. Pat. No. 6,227,809, issued May 8, 2001, is hereby incorporated by reference. These applications and patents may disclose aspects of structures, components and processes related to pumps and fluid analyzers.

SUMMARY

The present invention is a thermal pump for fluid analyzers.

BRIEF DESCRIPTION OF THE DRAWING

FIGS. 1 a and 1 b show a top cross-section view and a side cross-section view of an illustrative example of a thermal pump;

FIG. 2 shows a graph of a comparison of a fluid analyzer thermal conductivity relative to concentrator pulses for several gases;

FIG. 3 shows a table of a relationship of various parameters of a pump design and performance factors for various design variables;

FIG. 4 shows an illustrative example of a thermal pump;

FIG. 5 shows a diagram of a check valve for a thermal pump;

FIG. 6 shows another illustrative example of a thermal pump;

FIGS. 7 a and 7 b are top and side views another illustrative example of a thermal pump;

FIG. 7 c is a perspective view of a check valve for a thermal pump;

FIG. 8 a shows a graph of a flow pulse from fluid analyzer elements relative to an input pulse to the separator elements;

FIGS. 8 b and 8 c show perspective and side views of illustrative example of a check valve 64;

FIG. 8 d shows a test set up for a check valve using a fluid analyzer and a flow sensor;

FIG. 8 e shows a schematic of a series of separator elements used for generating a pulse relevant to the graph in FIG. 8 a;

FIG. 8 f shows a graph of delta pressure, Δp, in bar on the ordinate coordinate versus a flow rate for a sample in a fluid analyzer separator chip, for various numbers of elements;

FIG. 9 shows a layout of an illustrative example of a thermal pump, without mechanical valves, situated in a fluid analyzer;

FIG. 10 is an illustrative timing diagram of activity related to the pump shown in FIG. 9;

FIG. 11 shows another illustrative example of a thermal pump, without mechanical valves, situated in a fluid analyzer; and

FIGS. 12-15 show an illustrative example of a fluid analyzer that may be used in conjunction with thermal pumps.

DESCRIPTION

Initially, injecting analytes from a concentrated air sample into a non-air carrier gas stream for further analysis may run the risk of losing “focus” (i.e., diluting the sample), and require the addition of hardware but could result in such things as slow micro-valves, and still not eliminate air components that render downstream detectors less sensitive. In the case of a H₂carrier gas from a generator, it may be especially difficult to balance the H₂flow to meet the need for exact gas velocity of the preconcentrator and for velocity steadiness of the separator for an appropriate analysis. To meet such requirements, one may need a good steady and adjustable pump which the present invention may provide.
FIGS. 1 a and 1 b show a top cross-section view and a side cross-section view of a thermal pump 10. Pump 10 does not necessarily have moving parts except for membrane-based check valves 11 and 12. The pump may be driven by short voltage and approximately 200° C. pulses (1-10 ms) to microbridge-type heating elements of heater 13 which induce rapid gas heating, expansion and flow as directed by the check valves 11 and 12. These micro-pump elements 11 and 12 may be stacked in series and parallel, to meet volume and pressure head requirements, on one wafer. A principle of the pump 10 is that it uses cyclical thermal expansion and contraction of the gas to be transported with operating two passive check valves 11 and 12.
As the internal channel area 14 cools down, the channel pressure may decrease to a level less than the pressure at the inlet 15 and cause valve 11 to open and let a sample 16 enter the channel or chamber 14. When the pressures approach equal levels, then valve 11 may close. At this time, the heater 13 may be energized and heat up the sample 16 which expands and increases the pressure in chamber 14 to be greater than the pressure at an outlet 17. This may cause valve 12 to open and let the sample 16 to exit from the channel 14 via open valve 12 and outlet 17. With the heater 13 off, the sample 16 in the channel 14 may cool down and contract. This may cause the pressure in the channel 14 to be less than the pressure at the inlet 15 and thus causing valve 11 to open and let more sample 16 into channel 14. The heater may be energized again to heat up a new portion of sample 16 which expands and causes the pressure in channel 14 to increase. The pressure of channel 14 may exceed the pressure at outlet 17 thereby causing valve 12 to open and let some of the sample 16 to exit the channel 14 via valve 12 and outlet 17. Then, the heater or elements 13 may be de-energized resulting in the sample 16 in the channel 14 to contract and decrease the pressure in the channel 14. The cycle may continue to repeat as long as operation of pump 10 is desired.
The structure of pump 10 may have three main parts fabricated with MEMS (micromachined electromechanical systems) technology. The three main parts may be a bottom wafer 21, membrane 22, and top wafer 23. The first part may be the bottom channel wafer 21. The wafer 21 may have a recess 24 for an inlet and a recess 25 for channel 14. On the remaining portions of wafer 21, which may be the ridges 104 remaining after removal of the material for recesses 24 and 25, may be placed a membrane 22. Membrane 22 may have perforations 26 between the recess 24 of inlet 15 and channel 14, between a recess 27 and recess 25 of channel 14, and between recess 25 and a recess 28 at the outlet 17. Membrane 22 may have heater elements 13 in the middle set of perforations 26 between recesses 27 and 25. Current to drive the heater elements 13 may come in at terminals 29. Situated on membrane 22 may be remaining portions or ridges 105 of the top channel wafer 23. Wafer 23 may have recess 27 for a portion of channel 14 and recess 28 for outlet 17.
A characteristic of the pump 10 may reside in its micro-fluidic scale (facilitating a high cycling frequency), the use of low-mass, low-voltage, thin-film heater or elements 13 (for high efficiency) and a low-cost design that lends itself to monolithic integration into a micro-fluidic MEMS structure. The pump 10 may be used in conjunction with a phased heater array for enhanced detection (PHASED), which is described herein. The PHASED system may be implemented with a fluid analyzer or specifically a micro gas analyzer (MGA). It appears that the membrane-supported, low-mass, and thermally isolated heaters of the PHASED MGA structure, might lend themselves to the present gas pumping approach.
One may observe that gas expansion and contraction pulses of these heaters may generate signals in TCDs (thermal conductivity detectors) which sense such pulses as convective heat transfer changes, as shown in a graph of FIG. 2. This graph reveals a comparison of PHASED temperature conductivity (TC) to pre-concentrator (PC) pulses in three sample gases, at a flow of about 120 cm/s (at 3 psid). The graph shows normalized TC difference in bits versus time from the last PC pulse in the milliseconds (ms). The indicated pulses may be of an actual gas-expansion nature rather than an electronic pick-up nature that could be proven when such pulses are measured with gases of different thermal conductivities, but for the same excitation voltage pulses to the heaters. One may interpret the much reduced pulses shown for He, relative to those measured with N₂and Ar, as being caused by the much higher (about 6-8 times) thermal conductivity of He, relative to that of N₂and Ar, respectively.
The principle of operation may involve a voltage pulse to a thermally isolated membrane heater in a gas volume, V, to raise its temperature by ΔT=200° C., heat and expand the surrounding gas by a maximum factor of ΔV_h/V=(273+200)/(273+25)=1.587, or 58.7%. The actual expansion may be less because only about ⅔ of the gas reaches that temperature maximum.
During cool down, the gas may maximally contract from its ideal maximum down to a fraction of ΔV_c/V=(273+25)/(273+200)=0.598 of the hot and expanded volume. One may take a more conservative number of 40% for expansion and contraction of the gas volume around one heater of V=0.01·0.01·0.5=0.00005 cm³, but operate it at a frequency of f=830 Hz. By constraining or channeling this expansion and contraction action of the gas with check valves 11 and 12 as depicted in FIG. 1, one may induce a flow past the check-valves that might generate a flow of f·ΔV=830·0.00005·0.4·60˜1 cm³/min, if dead volumes have no effect on the total amount of flow so generated. Of course, one might arrive at that volumetric flow at a lower frequency, if more heater elements 14 are used in parallel.
The efficiency of such a pump 10 may be compared with that of other pumping approaches. If one defines efficiency as the ratio η=Output/Input=(ideal compression work on the gas)/(actual energy), one may identify for a gas volume V=50 nL=5·10⁻⁵cm³as one has in each chamber surrounding one of the PHASED heater array elements with the following factors. Ideal expansion output energy of each stroke, may be E_out=p·ΔV/(1+V_Dead/V)=10⁶·0.58·5·10⁻⁵/(1+0.2)=24.17 erg, assuming a 20% dead volume, which reduces the effect of the 58% expansion. Input thermal energy into gas, may be E_in,gas=n·c_p·ΔT=V/V_M·c_p·ΔT=5·10⁻⁵/22500·7·4.184·10⁻⁷7·200=130.2 erg. The input energy into a membrane heater, steady-state, using x=430° C./W, may be E_in,htr,ss=Δt·ΔT/x=0.002·200/430=9302 erg, for each heater pulse of 2-ms duration. With the input energy into a membrane heater, transient, using ρ(Si₃N₄)=3.2 g/cm³and c_p=0.71-0.9 J/g/° C., one may get E_in,htr,tr=c_p·ΔT·ρ·V=0.8·180·3.2·0.5·0.01·0.0001=0.23 mJ=2300 erg. This may correspond to an efficiency, η=Eout/Ein, approaching 0.2%. This would be based on the measured, presently used but still very lossy heater, which may be attached continuously to the Si substrate. An improved design ( 1/10^thattachment to a polymer) may reduce this loss by ≦10 times, so that the efficiency might approach 2%. Further efficiency improvements may be achievable with shorter pulses, e.g., Δt=0.5 to 1 ms instead of 2 ms as used herein (but long enough to expel the heated gas), keeping in mind that x=430° C./W was measured under steady state conditions. The noted 2% would only be about five times lower than computations for other micropumps. Sustaining the energy input at or above 830 Hz rate could release E_in,htr·f=9302/10⁷0.830=0.772 watts in the worst case, which may be too high for portable devices, but suitable for analyzers hooked up to line power.
FIG. 3 shows a table that may be used in a pump stack design, with assuming isothermal compression. The table provides a relationship of various parameters of the pump design and its performance factors for various design variables.
Relative to making a pump, FIG. 4 shows another illustrative example of a thermal pump 20. Having a PHASED chip available, it should be easy to add a cover plate 31 as shown, situated on a wafer 35, which would not only seal any leaky membrane gas heaters 32, but also host one of the indicated polymer-film membranes 33 and hold, e.g., a capillary 34 at the outlet, which may lead to a flow and/or pressure sensor to quantify the pump's performance. By energizing a rising number of heating elements 32 (up to 100) in parallel, one may be able to quantify the decreasing influence of dead volume. Increasing frequency, from a low 1-10 Hz to 1000 Hz, may quantify the maximum rate and associated efficiency of pump 20.
Heater membrane 33 may be formed on a surface of heater wafer 35. Heater 35 may be situated on a channel wafer 36 with the membrane situated between wafers 35 and 36. A sample may be drawn in through a port 37 and a polymer check valve 38 into a channel 39 of wafer 33 via a port through heater wafer 35. Check valve 38 may be attached to cover plate wafer 31. The sample, along with thermal pump operational principles described herein for pump 10, may be moved through channel 39, past heater 32, and out of channel 39 through a port in heater wafer 35 and a check valve 41. Check valve 41 may be attached to the top of heater wafer 35 over the port leading from the chamber 39. The sample may go through a port 42 into a capillary 34 connected to a flow and/or pressure sensor. Leads 45 may provide electrical power for the heater 32.
FIG. 5 shows an example of a MEMS check valve 38 or 41 that may be placed over the opening of the cover plate 31 or heater wafer 35, respectively. The valve may have a hinge 43 at the left and guides 44 at the right. The check valves 38 and 41 may be fabricated by staking a thin-film polymer film on 2-3 points around a respective PHASED inlet or outlet orifice.
An illustrative example thermal pump 30 in FIG. 6 shows that one may also achieve the pumping action in one plane, i.e., without requiring the gases to move from one level to another, which can have advantages in reducing viscous friction losses and in making the assembly more compact. Pump 30 may be similar to pump 20 except that pump 30 does not have the cover plate 31 or the check valves 38 and 41 of pump 20. The sample may enter a port 46 of heater wafer 35 and enter channel 39 of wafer 36 via a polymer check valve 47. Through the operation of heater 32 and thermal pump principles of pump 10, the sample may exit chamber 39 via a polymer check valve 48 and a port 49 in heater wafer 35 to capillary 34. Capillary 34 may be coupled to flow and/or pressure sensors. Heater 32 may be situated in a membrane 33 and be powered via terminals 45.
A curious effect of steam power is that it seems to be effective on a small scale. If one could keep steam from condensing, one may push a “plug” of gas at the needed PHASED flow rate of 0.5 cm³/min by only evaporating about 6 μg/s, which should only require about 15 mW.
The present thermal pump 30 may have the following characteristics over other designs for 0.1-1 cm³/min pumps. The present pump may harness a gas property (thermal expansion) for service as a micropump under advantageous micro “surface-to-volume” ratios which enable a much higher frequency operation than larger, conventional size pumps which have much smaller and thus impractical surface/volume ratios, despite being thermodynamically similar to combustion and Carnot cycle engines. The present pump may be integrated into a MEMS micro-fluidic structure and be fabricated at the wafer-level. Fabrication of the pump may be quite simple and not be dependent on critical film material properties. The pump may operate with low-power and low-voltage. It may generate gas pressure to drive a liquid, specifically for compact, portable dosimetry and microfluidic applications.
The advantages of the illustrative examples of thermal pumps 10, 20 and 30 over other possible pumping approaches may include the following items. The pump may be more easily integrated into a MEMS micro-fluidic structure than electrostatic pumps (EPs). It may be more compact than other previously conceived pumps. The present pump may be of lower-cost but may have some higher power consumption than EPs. The pump may have a lower drive voltage than EPs.
Another illustrative example is of a thermal pump 40, shown in FIGS. 7 a and 7 b, which may just have moving parts for membrane-based check valves 51 and 52. FIGS. 7 a and 7 b are top and side cross-section views of pump 40, respectively. The operation of pump 40 may be driven by the short voltage pulses to microbridge-type heating elements 53. These elements 53 may cause rapid (1-10 ms) gas heating to ≧200° C., which can induce gas expansion and flow as directed by the check valves 51 and 52, according to the thermal pump principles of pump 10. These check valves may have fluid-dynamic diode devices or reed-valve devices. This pump 40 may overcome previous fabrication problems by resorting to a three-wafer PHASED structure. A number of pumps 40 may be stacked as micro elements in series and/or in parallel to meet volume and pressure head requirements, on one wafer, or be distributed along a gas chromatography (GC) separator.
In pump 40, there may be a bottom wafer 54 having a heater and check valve wafer 55 situated on the wafer 54. On wafer 55 may be situated a top wafer 56. A recess in the heater wafer 55 may be a chamber or channel 57 for heating a sample with a heater having elements 53. The chamber may, for example, have about 100 nL of volume with approximate dimensions of about 0.1 mm×0.5 mm×1 mm×2, for above and below the heater elements 53. The sample may enter a port 58 from the bottom of wafer 54 into a space provided by the heater wafer 55 as a spacer between wafers 54 and 56. The sample may proceed through valve 51, past heater elements 53 in channel 57, and through valve 52. The sample may exit through a port 59. The heater elements 53 may be of a thin-film Pt heater. The movement of the sample though pump 40 may be based on the principles of operation of pump 10 as described herein.
The valves 51 and 52 may each be, for example, an array of nine 130 by 100 micron flaps situated over 50 micron holes. A flap 61 is shown in FIG. 7 c. Instead of reed-type like the flaps, the valves 51 and 52 may be a fluid dynamic type.
The pump 40 may use cyclical thermal expansion and contraction of the sample gas to be transported in a sensor by the operating two passive check valves, which may be reed-type or fluid-dynamic-type. Characteristics of this pump may include its micro-fluidic scale (facilitating high cycling frequency), the use of low-mass, low-voltage, thin-film heaters 53 (for high efficiency) and a design that lends itself to monolithic integration into a micro-fluidic MEMS structure. Integrating several of these pumps 40 along the separator (and even have the heaters support a stationary phase), opens up the prospect of overcoming one present “pressure drop” limit to microminiaturization, as the channels become ever smaller. Pump 40 may provide a low-cost and manufacturable gas sample pumping that can be integrated into micro-fluidic MEMS or PHASED structures.
It may be noted that the membrane-supported, low-mass, and thermally isolated heaters 53 of a PHASED MGA structure, might lend themselves to a different gas pumping approach. The pump 40 may be such that no pump cavity wall needs to be sealed against a flexible membrane. It may increase the efficiency by at least two times by allowing a sample gas to contact the heater elements 53 from top and bottom sides, as is revealed by cavity or channel 57 in FIGS. 7 a and 7 b due to the three-wafer structure. The pump's workability may be evidenced by measuring the flow pulse 62 resulting from one heater pulse 63 as shown in FIG. 8 a. FIG. 8 a shows a graph of one thermal pulse from 50 separator elements with a 12 volt input heater pulse 63 to the separator elements of a PHASED chip generating one flow pulse 62 of 1.2 cm³per minute using one check valve. The graph shows V, a flow rate in cm³/min on the left ordinate axis and E, input in volts to the heater elements of the PHASED chip, on the right ordinate axis, versus z, time in milliseconds, on the abscissa axis.
FIG. 8 a shows that the measurement with 50 elements heated to only about 155° C. generated a flow pulse such that an estimated 16 elements could be operated to yield a flow of 0.4 cm3 and consume only 0.48 W. FIGS. 8 b and 8 c show the check valve 64 with an illustration and a cross-section view, respectively. FIG. 8 d shows a test set up with the check valve 64, PHASED chip 65 and flow sensor 66, which may be used to obtain data like that plotted in the graph of FIG. 8 a. FIG. 8 e shows a schematic of the PHASED chip 65 showing a series of the 50 separator elements for generating a thermal pulse relevant to the graph in FIG. 8 a. FIG. 8 f shows a graph of delta pressure, Δp, in bar, on the ordinate coordinate, versus V, a flow rate in cm³/min, for a sample in the PHASED separator chip 65, for 1, 2, 4 and 8 elements, respectively.
In summary, the thermal pump 40 over other designs for 0.1-1 cm³/min pumps may involve the following items. The pump 40 may harness a gas property (thermal expansion) for service as a micropump under advantageous micro “surface-to-volume” ratios which can enable much higher frequency operation than larger, conventional size pumps which have much smaller and thus impractical surface/volume ratios, despite thermodynamically being similar to combustion and Carnot cycle engines. This pump 40 may be integrated into a MEMS micro-fluidic structure and be completely fabricated at the wafer-level. The pump 40 may require just very simple fabrication and not be dependent on critical film material properties, such as an EP. It may operate with low-power at low-voltage. Pump 40 may generate gas pressure to drive a liquid, specifically for compact, portable dosimetry and microfluidic applications.
Pump 40 may have good manufacturability. This thermal pump may be distributed as elements along a GC separator. Such pump elements may do double duty by serving as coated GC separator elements of a PHASED MGA, despite their pulsations. The MGA may save significant energy by the elements doing such double duty.
In the lower left corner of FIG. 9 is a layout of a valveless thermal pump 50. The pump 50 may operate with a low drive voltage, and have no moving parts, mechanical check valves or actuators. This pump 50 may be fabricated via standard micromachining processes, and thus be fully integratable—monolithically—with microsensor(s) at a wafer level. Its fabrication costs may be low, and because of a lack of moving parts, and it may have a very long service life. Smart duty-cycling, the small size of the pump 50, and a low-heat dissipation design may mitigate the power consumption of the electrically pulsed heaters in the pump.
The thermal pump 50 of FIG. 9 may be in a context of a PHASED MGA 60 having a pre-concentrator (PC) 101 and a separator 102. For instance, there may be three heaters 72, 73 and 74 situated in the PC 101 channel. A sample may enter the PC 101 at an inlet 75. After proceeding through the PC 101 and the separator 102, the sample may exit the separator 102 at an outlet 76. The electrical aspects of PHASED set-up or analyzer 60 may include chemical impedance sensors (CIDs) 77 and 78 at inlet 75 and outlet 76, respectively, connected to an electronics (control) module 79. The CIDs 77 and 78 may be connected in a differential mode and provide signals with information about the sample to module 79. Module 79 may provide heater control for heaters of PC 101 and pump heaters 72, 73 and 74 of pump 50. Heater control may be provided to PC 101 via contact pads 81. Module 79 may contain a pre-amplifier, an analog-to-digital converter, a timer and a microprocessor. The microprocessor may provide PC 101 and separator 102 heater control signals, pump heater control signals, and TCD power. It may provide timed energizing and de-energizing signals to heaters 71, 72 and 73. This thermal pump 50 in pre-concentrator 101 may have instead, for example, just two heaters 72 and 73 operating with appropriate pulses for gas expansion and contraction. The design or programming of the sequences of the pulses may result in an effective thermal valve 103 between heaters 72 and 73, taking advantage of the viscosity changes to provide a variable viscous drag of changing density of the gas to result in a phenomenon of a thermal check valve, and so as to provide an effective pumping of the gas down a channel of the preconcentrator 101 which may have in some backward stepping and forward stepping of the gas in the channel, but with a net forward stepping of the gas. Heater elements of the preconcentrator may be selected as heaters 72 and 73 for the thermal pump 50. The pump 50 could be instead placed in separator 102 and use separator elements as heater 71 and 72, and also as heater 73 if desired for possibly more efficient pumping.
Module 79 may be programmed in various ways to provide effective pumping by thermal pump 50. The region of the thermal valve 103, for instance, between heaters 72 and 73 (just as in a three or more heater thermal pump 50) may be made more effective by making the region cross-section smaller and flow governing, and length short relative to the region's distance from the heaters, or the distance between adjacent heaters of the pump, to minimize energy losses. The region's shape may be designed to emulate a fluidic diode which can be heated to enhance its “closed” period effectiveness, and/or have a cross-section shrink which is a coating with a film that expands with temperature. The coating may be on the channel wall at the region of the thermal valve 103 which may upon expansion, due to temperature change, narrow the passage of the channel for fluid. Also, electronics module 79 may be connected to thermal conductivity detectors (TCDs) situated in analyzer 60 as needed.
The pump 50 may use reciprocating thermal expansion and contraction of the same sample gas to be transported to a sensor or device. This pump 50 may leverage temperature-dependent, variable gas viscosity in a first approach, augmented by a fluidic diode effect in a second approach, and by thermal expansion films to pinch off flow in a third approach as fixed fluidic (leaky) check valves.
In addition to the characteristics of the thermal micro-pump 10 with its micro-fluidic scale (facilitating high cycling frequency), the pump 50 may likewise use low-mass, low-voltage, thin-film heaters (for high efficiency), and with a design that may lend itself to monolithic integration into a micro-fluidic MEMS structure. However, the pump 50 does not necessarily require the mechanical check valves, 11 and 12, 38 and 41, 47 and 48, or 51 and 52, as pumps 10, 20, 30 and 40 may need, respectively. Because of a lack of mechanical valves, pump 50 may provide very low-cost and long-life as a gas sampling pump which may be integrated into micro-fluidic MEMS structures.
It seems that the membrane-supported, low-mass, and thermally isolated heaters of the PHASED MGA structure, may lend themselves to a gas pumping scheme of pump 50. One might observe that gas expansion and contraction pulses of the heaters in the MGA structure may generate signals in TCDs (thermal conductivity detectors), which sense such pulses as convective heat transfer changes, as shown in FIG. 2. That the indicated pulses are of an actual gas-expansion nature rather than an electronic pick-up one may be shown upon a measuring of such pulses with gases of different thermal conductivities, but for the same excitation voltage pulses to the heaters. One could interpret the much reduced pulses shown for He, relative to those measured with N₂and Ar, as being caused by the high (about 6-8 times higher) thermal conductivity of He (relative to those of N₂and Ar), which may strongly reduce the gas temperature rise, before the heat is conducted to the Si substrate.
As to the principle of operation of thermal pump 50, it may help to note that a voltage pulse to a thermally isolated membrane heater in a gas volume, V, to raise its temperature by ΔT=200° C., will heat and expand the surrounding gas by a maximum factor of ΔV_h/V=(273+200)/(273+25)=1.587, or 58.7%. However, the actual expansion may be less because only about ⅔ of the gas may reach that temperature maximum, or about 39%.
As to the basic pump 50 operation, the non-mechanical check valves (i.e., of valveless pumps) schematically indicated in FIG. 10 may operate on one of several approaches. The first approach of pump 50 in FIG. 9 may be of variable viscous drag of the thermal check-valve when containing heated gas, which consists of regular PHASED heater elements. This first approach may require at least two or more heaters. The more heaters involved, the larger are the flow and pressure head. The heaters' synchronized and pulsed heating may be akin to moving a gas volume element forward like an inch-worm, in the sense that the actuation of adjacent heaters may cause some “back-stepping” and somewhat larger “forward stepping”.
The chart of FIG. 10 is a timing and flow chart of one set of valveless pump 50 heater elements 72, 73 and 74 (in FIG. 9) using equal size elements for expansion and as check valves. Higher efficiency pumps 50 may be designed with smaller check valve (CV) elements. The chart reveals thermal pump parameters on the left ordinate coordinate showing gas volume 82 at an upstream valve, and volts 83 and temperature 84 for the upstream valve. This left coordinate also shows an expansion rate 85 (>0) and a contraction rate 86 (<0), and volts 87 for a main expander. This coordinate shows volts 88 and a temperature 89 for a downstream valve. Also, the coordinate shows an amount of main expander flow 91 in a desired positive manner at the main element, relative to a zero magnitude 92. For one cycle 93, it may be noted that the cumulative amount of flow indicated by a sum of the positive flow 91 and negative flow 94 (areas between the curves and the zero level) is a positive flow. On the ordinate coordinate on the right side is the position of the heater array elements 72, 73 and 74 relative to the upstream valve, main expander and the downstream valve, respectively, for a sample flow 95.
The second approach of pump 50, in the lower left part of FIG. 11 for system 60, may include fluidic diodes 96 and 97. For instance, reciprocating flow in small channels driven by mechanical, piezoelectric, or thermal effects may serve to ratchet fluid forward, when connected to fluidic diodes 96 and 97. The non-mechanical property of the “valves” 96 and 97 may be a reason to designate pump 50 in FIG. 11 as “valveless”. Pump 50 in FIG. 11 may have the upstream fluidic valve 96, a heater 98, and the fluidic valve 97, in that order, going downstream. This pump's operation may be analogous to that of the pump in FIG. 9.
During cool down, the gas volume may contract from its maximum down to a fraction of ΔV_c/V=(273+25)/(273+200)·(2/3)=0.39 of the hot and expanded state. This 39% expansion and contraction of the gas volume around one heater of V=0.01·0.01·0.5=0.00005 cm³, may operate at a frequency of f=830 Hz. By constraining or channeling this expansion and contraction action of the gas with non-mechanical check valves as depicted in FIGS. 9 and 10, one may induce a flow past the check-valves that may generate a flow of f·ΔV=830·0.00005·0.39·60 ˜1 cm³/min, if all the flow is in the forward direction. But as indicated in FIG. 10 by the shaded areas, some reverse flow may occur if these check valves are leaky, which should be compensated with more pumping (e.g., a larger pump and/or a higher frequency). However, non-mechanical check valves may have less dead volume than the mechanical check valves, including less inertia and no stiction problems. Of course, one might arrive at the needed volumetric flow at a lower frequency, if more such pump 50 assemblies were used in parallel.
One approach may consist of two pumps 50 in series, so that a “contraction period” of the upstream one may coincide with an expansion period of the downstream one, while the “valve” in between the two is “open”. Similarly, the effectiveness of this valve may be improved by making its cross section small and “flow-governing”, its length short to minimize energy losses and its shape to emulate fluidic diodes 96 and 97 as indicated in FIG. 11. This valve may be heated to enhance its “closed” period effectiveness and its cross section shrink due to a coating with a film that expands with temperature.
One may look at the above improvement possibilities from the ground up by first considering the basics. To start, as to viscosity temperature dependence, 66% of the maximum temperature and viscosity change would be equivalent to a valve flow ratio of η/η_o=(T/T_o)^0.7={(273+25+200(2/3))/(273+25)}^0.7=1.295. As to valve coating thermal expansion, the strong influence of a channel cross section on the laminar flow rate, V in cm³/s, may make for a potentially very effective contribution, since V=π·Δp·r⁴/(8·L·η). One may view r as a hydraulic radius, r=h/(1+1·h/w), with h=height and w=width. A rectangular channel beginning at h_o=w_o/2 and ending at h=w_o/10 would result in a drop in V of V/Vo=[{h/(1+1·h/w_o)}/{h_o/(1+1·h_o/w_o)}]⁴=(2/10)4·(1.5/1.1)4=0.0055, which may be regarded as favorable for control of a fluid.
One may seek a material with a needed thermal expansion. If one starts with a rectangular channel with h_o=w_o/2, it may be one of the following alternatives. One may be of thin film where an approximate s_o=1 μm film that expands 25 times to 25 μm in a channel with w_o=60 μm and h_o=30 (including the film at T_o) and decreases to h=6 μm. However, materials with a temperature coefficient of expansion (TCE)=24/200=120,000 ppm/° C. might not be available. Another alternative may be a thick film where s_o=25 μm film that expands about two times (TCE=5000 ppm/° C.) to 49 μM causing again a decrease to h=6 μm. Or one may have a channel with h_o=w_o/10 that narrows to h=w_o/100, where h_o=6 μm, h=0.6 μm, s_o=25 μm, s=30.4, which may have a TCE=30.6/25−1=1120 ppm/° C. (e.g., polyethylene is 120-200 ppm/° C., PDMS=310 ppm/° C., and plastic wood=1230 ppm/° C.).
One may calculate the time needed to heat up a 25 μm film of thermal conductivity of k=0.01 J/(s·cm·K) and specific heat, c_p=1.67 J/(g·K), ρ=1.14 g/cm³. Thus, ν=k/c_p/ρ=0.01/1.67/1.14=0.00525 cm²/s, which means that a temperature rise of the heater will propagate across a film thickness of s=0.0025 cm in a time, t=s²/(2ν)=0.0025²/2/0.00525=0.00060 s, i.e., fast enough.
One may compare the efficiency of such a pumping approach (or pump 50) with that of other approaches. If one defines efficiency as the ratio η=Output/Input=(ideal compression work on the gas)/(actual energy), one may identify for a gas volume V=50 nL=5·10⁻⁵cm³, in each chamber surrounding one of the PHASED heater array elements.
An ideal expansion output energy of each stroke may be E_out=p·ΔV/(1+V_Dead/V)=10⁶·0.39·5·10⁻⁵/(1+0.2)=16.3 erg, assuming a 20% dead volume, which reduces the effect of the 39% expansion. The input thermal energy into the gas may be E_in,gas=n·c_p·ΔT=V/V_M·c_p·ΔT=5·10⁻⁵/22500·7·4.184·10^7·120=78.1 erg, where ΔT=(200−20)(2/3)=120. This energy may be provided by the membrane heaters and should not be counted. The input energy into membrane heater, using x=430° C./W, may be E_in,htr=Δt·ΔT/x=0.002·180/430·10⁷=8370 erg, for each heater pulse of 2 ms duration. The inputs into two ˜0.5 mm “check valve” heaters, using x=430° C./W, may be E_in,htr=Δt·ΔT/x=0.002·180/430/5·10⁷=1674 erg, for heater pulses of 2 ms duration.
These parameters may correspond to an efficiency of 16.3/(8370+1674)=0.16%. During a proposed Δt=4 second analysis time and a flow of V=0.5 cm³/min (driven by a pump element volumetric displacement of V_p=0.39·50·10⁻⁶), which requires a frequency of f=V/V_p=0.5/60/(0.39·50·10⁻⁶)=427 Hz, the total pump energy per one analysis period may thus be E(pump)=E(stroke)·f·Δt=(8370+1674)·427·4·10⁻⁷7=1.71 J.
The latter approach may use the measured (430° C./W), still very lossy, long and narrow heater, which is attached continuously to the Si substrate. Another approach ( 1/10^thattachment to a polymer) may reduce this loss by 10 times, so that the efficiency might then increase to 1.6% and the energy needed per analysis drop to 0.171 J. Further efficiency improvements may be achievable with shorter pulses, e.g., Δt=0.5 to 1 ms instead of 2 ms, as used above, keeping in mind that x=430° C./W was measured under steady state conditions. The 1.6% would only be about six times lower than the latest computations for an EP stack. The 171 mJ amount of energy per analysis appears to compare favorably with earlier estimates.
Having PHASED chips available, in view of a FIG. 11 (or 9) illustration of an example, it may be easy to add a cover plate as shown, which would not only seal any leaky membrane gas heaters, but also host one of the indicated polymer-film membranes and hold, e.g., a capillary at the outlet, which may lead to a flow and/or pressure sensor to quantify the pump's performance. By energizing a rising number of elements (up to 100) in parallel, one may be able to quantify the decreasing influence of dead volume. Increasing frequency from a low 1-10 Hz to 1000 Hz may quantify the maximum rate and associated efficiency of such a pump.
For a demonstration, the check valves may be fabricated by staking a thin-film polymer film on 2-3 points around the PHASED outlet orifice.
The second approach in FIG. 10 shows that one may also achieve the pumping action in one plane, i.e., without requiring the gases to move from one level to another, which may have the advantages in reducing viscous friction losses and in making the assembly more compact.
In summary, the characteristics of the valve-less thermal pump 50 over other 0.1-1 cm³/min pumps may include the following items. The pump may have more reliable and efficient (higher on/off flow ratio) than micropumps dependent on mechanical check valves. Such pump may be less prone to failure caused by small particulate matter. The pump may harness well known gas properties (thermal expansion and temperature-dependence of viscosity) for service as a micropump under advantageous micro “surface-to-volume” ratios, which may enable much higher frequency operation than conventional size pumps, which would not even be practical, despite being thermodynamically similar to combustion and Carnot cycle engines. Pump 50 may be integrated into a MEMS micro-fluidic structure and be completely fabricated at the wafer-level. It may involve very simple fabrication. The illustrative examples (of FIGS. 9 and 10) are not necessarily dependent on critical film material properties. Another illustrative example may depend on the thermal expansion of the check-valve materials to increase the on/off flow ratio. Pump 50 may operate with low-power and low-voltage. Enhanced operation may be achieved by combining fluid-mechanical-inertial-Co and a effect valves with the noted herein thermal effects, i.e., heating such fluidic diodes during their “closed” or “reverse-flow” period to enhance their effectiveness. There may be enhanced operation by an addition of “thermal-expansion” thick-film segments, which can “pinch off” the flow when such segments (with high TCE-films) are heated. The check-valve segments may be merely 1/10^thor less in length, relative to the “gas expansion” segments. The “gas expansion” segments may be of a greater width, to minimize energy losses, resulting in an optimized design—minimum heat loss, yet enabling fast heating and cooling. One may implement a multi-stage operation to achieve greater pump pressure heads by connecting two of more sets of pumps in series.
A fluid analyzer which may be used in conjunction with the thermal pump may include a channel or channels for a flow of a sample along a membrane that supports heaters and a stationary phase for sample analysis. The channel or channels may be an integral part of the micro fluid analyzer. The analyzer may have the pre-concentrator (PC) 101 (viz., concentrator) and chromatographic separator (CS) 102 that incorporates the channel or channels. FIG. 12 is a system view of an example fluid analyzer which may be a phased heater array structure for enhanced detection (PHASED) micro gas analyzer (MGA) 110. It reveals certain details of the micro gas apparatus 110 which may encompass the specially designed channel described herein. The PHASED MGA 110, and variants of it, may be used for various fluid chromatography applications.
Sample stream 111 may enter input port 112 to the first leg of a differential thermal-conductivity detector (TCD) (or other device) 115. A pump 116 may effect a flow of fluid 111 through the apparatus 110 via tube 117, though pump 116 may be a thermal pump or be replaced by a thermal pump. There may be additional pumps, and various tube or plumbing arrangements or configurations for system 110 in FIG. 12. Fluid 111 may be moved through a TCD 115, concentrator 101, flow sensor 122, separator 102 and TCD 118. Controller 119 may manage the fluid flow, and the activities of concentrator 101 and separator 102. Controller 119 may be connected to TCD 115, concentrator 101, flow sensor 122, separator 102, TCD 118, and pump 116. The pump 116 may be a thermal pump or be replaced with a thermal pump integrated in the concentrator 101 or separator 102. Data from detectors 115 and 118, and sensor 122 may be sent to controller 119, which in turn may process the data. The term “fluid” may refer to a gas or a liquid, or both.
FIG. 13 is a schematic diagram of part of the sensor apparatus 110 representing a portion of concentrator 101 and/or separator 102 in FIG. 12. This part of sensor apparatus 110 may include a substrate or holder 124 and controller 119. Controller 119 may or may not be incorporated into substrate 124. Substrate 124 may have a number of thin film heater elements 125, 126, 127, and 128 positioned thereon. While only four heater elements are shown, any number of heater elements may be provided, for instance, between two and one thousand, but typically in the 20-100 range. Heater elements 125, 126, 127, and 128 may be fabricated of any suitable electrical conductor, stable metal, alloy film, or other material. Heater elements 125, 126, 127, and 128 may be provided on a thin, low-thermal mass, low-in-plane thermal conduction, membrane or support member 124, as shown in FIGS. 13 and 14.
Substrate 130 may have a well-defined single-channel phased heater mechanism 131 having a channel 132 for receiving the sample fluid stream 111, as shown in FIG. 14. The channels may be fabricated by selectively etching silicon channel wafer substrate 130 near support member 124. The channel may include an entry port 133 and an exhaust port 134.
The sensor apparatus 110 may also include a number of interactive elements inside channel 132 so that they are exposed to the streaming sample fluid 111. Each of the interactive elements may be positioned adjacent, i.e., for closest possible contact, to a corresponding heater element. For example, in FIG. 14, interactive elements 135, 136, 137, and 138 may be provided on a surface of support member 124 in channel 132, and be adjacent to heater elements 125, 126, 127, and 128, respectively. There may be other channels with additional interactive film elements which are not shown in the present illustrative example. The interactive elements may be formed from any number of films commonly used in liquid or gas chromatography. Furthermore, the above interactive substances may be modified by suitable dopants to achieve varying degrees of polarity and/or hydrophobicity, to achieve optimal adsorption and/or separation of targeted analytes.
Controller 119 may be electrically connected to each of the heater elements 125, 126, 127, 128, and detectors 115 and 118 as shown in FIG. 13. Controller 119 may energize heater elements 125, 126, 127 and 128 in a time phased sequence (see bottom of FIG. 15) such that each of the corresponding interactive elements 135, 136, 137, and 138 become heated and desorb selected constituents into a streaming sample fluid 111 at about the time when an upstream concentration pulse, produced by one or more upstream interactive elements, reaches the interactive element. Any number of interactive elements may be used to achieve the desired concentration of constituent gases in the concentration pulse. The resulting concentration pulse may be provided to detector 118, for detection and analysis.
FIG. 15 is a graph showing illustrative relative heater temperatures, along with corresponding concentration pulses produced at each heater element. As indicated above, controller 119 may energize heater elements 125, 126, 127 and 128 in a time phased sequence with voltage signals 150. Time phased heater relative temperatures for heater elements 125, 126, 127, and 128 may be shown by temperature profiles or lines 151, 152, 153, and 154, respectively.
In the example shown, controller 119 (FIG. 13) may first energize first heater element 125 to increase its temperature as shown at line 151 of FIG. 15. Since first heater element 125 is thermally coupled to first interactive element 135 (FIG. 14), the first interactive element desorbs selected constituents into the streaming sample fluid 111 to produce a first concentration pulse 161 (FIG. 15) at the heater element 125, if no other heater elements were to be pulsed. The streaming sample fluid 111 carries the first concentration pulse 161 downstream toward second heater element 126, as shown by arrow 162.
Controller 119 may next energize second heater element 126 to increase its temperature as shown at line 152, starting at or before the energy pulse on element 125 has been stopped. Since second heater element 126 is thermally coupled to second interactive element 136, the second interactive element also desorbs selected constituents into streaming sample fluid 111 to produce a second concentration pulse. Controller 119 may energize second heater element 126 such that the second concentration pulse substantially overlaps first concentration pulse 161 to produce a higher concentration pulse 163, as shown in FIG. 15. The streaming sample fluid 111 may carry the larger concentration pulse 163 downstream toward third heater element 127, as shown by arrow 164.
Controller 119 may then energize third heater element 127 to increase its temperature as shown at line 153 in FIG. 15. Since third heater element 127 is thermally coupled to third interactive element 137, third interactive element 137 may desorb selected constituents into the streaming sample fluid to produce a third concentration pulse. Controller 119 may energize third heater element 127 such that the third concentration pulse substantially overlaps larger concentration pulse 163 provided by first and second heater elements 125 and 126 to produce an even larger concentration pulse 165. The streaming sample fluid 111 carries this larger concentration pulse 165 downstream toward an “Nth” heater element 128, as shown by arrow 166.
Controller 119 may then energize “N-th” heater element 128 to increase its temperature as shown at line 154. Since “N-th” heater element 128 is thermally coupled to an “N-th” interactive element 138, “N-th” interactive element 138 may desorb selected constituents into streaming sample fluid 111 to produce an “N-th” concentration pulse. Controller 119 may energize “N-th” heater element 128 such that the “N-th” concentration pulse substantially overlaps larger concentration pulse 165 provided by the previous N−1 interactive elements. The streaming sample fluid may carry the resultant “N-th” concentration pulse 167 to either a separator 102 or a detector 118.
In the present specification, some of the matter may be of a hypothetical or prophetic nature although stated in another manner or tense.
Although the invention has been described with respect to at least one illustrative example, many variations and modifications will become apparent to those skilled in the art upon reading the present specification. It is therefore the intention that the appended claims be interpreted as broadly as possible in view of the prior art to include all such variations and modifications.

Claims

1-10. (canceled)

11. A valveless thermal pump comprising:

a channel having an inlet and an outlet;

a first heater situated in the channel;

a second heater situated in the channel downstream from the first heater;

a thermal control region situated in the channel between the first heater and the second heater; and

a controller connected to the first heater and the second heater to provide a sequence of heater energizing signals to the first and second heaters to provide a reciprocating expansion and contraction of a fluid to result in the pumping of the fluid in the channel; and

wherein the thermal control region biases the fluid downstream.

12. The thermal pump of claim 11, wherein:

the thermal control region comprises an improvement; and

the improvement comprises:

a reduced cross section in the channel at a location of the thermal control region; and/or

a short length relative to a distance between the first and second heaters.

13. The thermal pump of claim 11, wherein the thermal control region comprises a film that reduces an area of a passageway of the channel upon a change of temperature of the film.

14. The thermal pump of claim 11, further comprising at least another heater in the channel proximate to the first and second heaters.

15. The thermal pump of claim 11, wherein the channel is a part of a fluid analyzer.

16. A thermal pump comprising:

a channel;

a first fluidic diode structure situated in the channel;

a second fluidic diode structure situated in the channel; and

a heater situated in the channel proximate to and between the first fluidic diode structure and the second fluidic diode structure.

17. The pump of claim 16, further comprising a controller connected to the heater for energizing the heater according to a sequence of signals for pumping a fluid through the channel.

18. The pump of claim 16 further comprising:

a second heater proximately associated with the first fluidic diode structure; and

a third heater proximately associated with the second fluidic diode structure; and

wherein the first and second heaters are energized at certain times to increase the effectiveness of the first and second fluidic diode structures.

19. The pump of claim 16, wherein the channel is a part of a fluid analyzer.

20. The pump of claim 19, wherein the fluid analyzer is a phased heater array for enhanced detection (PHASED) fluid analyzer.

21. A valveless thermal pump comprising:

a channel having an inlet and an outlet;

a first heater situated in the channel;

a second heater situated in the channel downstream from the first heater;

wherein the expansion and contraction of the fluid moves the fluid in downstream direction and substantially prevents the fluid from moving in an upstream direction without moving parts in the thermal control region.

22. The thermal pump of claim 21, wherein:

the thermal control region comprises an improvement; and

the improvement comprises:

a short length relative to a distance between the first and second heaters.

23. The thermal pump of claim 21, wherein the thermal control region comprises a film that reduces an area of a passageway of the channel upon a change of temperature of the film.

24. The thermal pump of claim 21, further comprising at least another heater in the channel proximate to the first and second heaters.

25. The thermal pump of claim 21, wherein the channel is a part of a fluid analyzer.