Micromachined flow sensors—a review

doi:10.1016/S0955-5986(97)00019-8

Flow Measurement and Instrumentation

Volume 8, Issue 1, March 1997, Pages 7-16

https://doi.org/10.1016/S0955-5986(97)00019-8 Get rights and content

Abstract

Micromachining technology has been developed very rapidly in recent years. This technology takes advantage of the benefits of semiconductor technology to address the manufacturing and performance requirements of the sensors industry. The compatibility of micromachining and microelectronics makes the integration of electronics and mechanical elements possible. This covers the need of low-cost, accurate and reliable sensors for industrial and consumer product applications. An important product of micromachining technology is the micro-mass flow sensor which has a history of over 20 yrs. This paper presents a review of the research and development of micromachined flow sensors which have been done in the last few years by international academic and industrial institutions.

Introduction

In the past 20 yrs, the application of microelectronic technology to the fabrication of mechanical, thermal and optical devices has greatly stimulated research and development of micromachined sensors, generally, and more specifically the silicon sensor. Because sensors for different physical variables are structurally different, there is no single technology that allows the fabrication of a wide variety of sensors and integrated electronics. Currently, there are four major classifications of micromachining technology: bulk-micromachining, surface-micromachining, epi-micromachining (or surface-proximity-micromachining) and LIGA-micromachining (LIGA: german acronyms for Lithographie-Galvanoformung-Abformung) 1, 2, 3. The last one is because of its incompatibility with microelectronics not often used for the fabrication of micro sensors.

In bulk-micromachining, the sensors are shaped by etching a large single-crystal substrate. A high-resolution etch and tight dimensional control are provided using anisotropic etching techniques. Surface-micromachined sensors are constructed entirely from thin films. Free standing and movable parts can be fabricated using sacrificial etching. Single-crystal materials used in bulk-micromachining have well-defined properties in contrast to those of amorphous polycrystalline thin films, hence yielding sensors with reproducible characteristics. The bulk-micromachining has the disadvantage that the devices are generally relatively large and therefore consume most of the chip area. A recent development which has tried to take the advantages from both technologies while minimising the disadvantages, is epi-micromachining. Epi-micromachining is essentially front-side bulk micromachining, using wet or dry etching, where the epitaxial layer forms the mechanical structure.

The flow measurement is a classical field of measurement technology. The working principles contain nearly all domains of physics (Fig. 1) [4]. The fast development of micromachining technology makes the realisation of conventional measurement principles possible. This opens a new market for new applications and products.

In the year 2000, flow measurement and control with micromachined flow sensors will share about 19% of the MEMS-market (MEMS: micro electromechanical systems) which will amount in all probability to US$ 14 billion [5]. The growing market of micromachined flow sensors requires a lot of research and development of this new technology and its use for fabrication of flow sensors. Governed by their small geometry, the biggest advantages of micromachined flow sensors are the low energy consumption and the possibility to measure very small mass flow (micro litres per minute).

The first flow sensor based on silicon technology was presented in 1974 by van Putten and Middelhoek [6]. In the 1980s when the micromachining and micro-electromechanical systems were established as common expressions in the professional world, some industrial firms developed the new product `micromachined flow sensor' based on the hot-film principle. In this type of flow sensor, the fluid streams around the sensor.

From the end of the eighties to the beginning of the nineties, the development of micromachined sensors has been an important research field of numerous international academic institutions. In the 1990s, the development tends towards the fabrication of complex micro fluidic systems (micro flow sensors, pumps and valves in a system). Therefore, there is a need for a flow sensor that can measure very small flow rates. The result of this challenge is a new class of micromachined flow sensors that has an integrated micro channel. In this type of micromachined flow sensor, the fluid streams inside the sensor. The first sensor of this type was presented by Petersen in 1985 [7].

Because of the high dynamics resulting from the miniaturisation, there are new interesting applications for micromachined flow sensors. An example is the thermal microphone of the MESA-institute (University Twente, Netherlands) that is able to measure acoustic flow [8].

Because of the enormous number of already developed micromachined flow sensors which are based on the thermal principle, there are two classifications: non-thermal and thermal flow sensors.

Section snippets

Non-thermal flow sensors

All the already realised non-thermal flow sensors are based on the mechanical working principle. For example, the flow can be measured indirectly by the drag force using a silicon cantilever. In laminar flow conditions, with a small Reynolds number, the drag force parallel to the flow direction is given by the Navier–Stokes-law: $F=CLvη,$ where F is the drag force, C a constant depending on the form of the cantilever, L the dimension of the cantilever, v the flow velocity and η the dynamic

Thermal flow sensors

International research works on the field of micromachined flow sensors showed that because of their structural and electronic simplicity thermal sensors can be realised diversely and easily using the micromachining technology. Fig. 5 shows the general signal transport of a thermal flow sensors.

With two heater control modes and two evaluation modes, there are six operational modes shown in Table 1 and three types of thermal flow sensors:

•
Thermal mass flow sensors which measure the effect of the

Applications of micromachined flow sensors

Micromachined flow sensors described in the last two sections cover a wide range of medium, flow range and applications from macroscopic air velocity measurements to the measurement, as well as control, of microlitres-per-minute of liquid flow.

Thermal flow sensors are suitable for the measurement of gas flow velocity. The sensor chip is located directly in the gas flow. Table 3 shows a survey of the performance of the above described sensors. With the typical square-root characteristic (Fig. 21

Development trend and outlook

The development of complex microfluidic systems consisting of different microfluidic components requires integrated flow sensors which are able to measure very small flow rates. The micro-dosing system is an interesting product for medicine, micro chemistry and micro biology.

The reduction of manufacturing costs is an important aspect for the industrial use of micromachined flow sensors. Monolithic integration of electronics on the sensor chip makes this possible.

The use of an integrated

References (60)

R. Kersjes et al.
A integrated sensor for invasive blood-velocity measurement
Sensors and Actuators A
(1993)
S.T. Cho et al.
A high-performance microflowmeter with built-in self test
Sensors and Actuators A
(1993)
P. Enoksson et al.
Fluid density sensor based on resonance vibration
Sensors and Actuators A
(1995)
J. Vollmer et al.
Bistable fluidic elements in LIGA technique for flow control in fluidic microactuators
Sensors and Actuators A
(1994)
A.F.P. van Putten
An integrated silicon double bridge anemometer
Sensors and Actuators
(1983)
Y.C. Tai et al.
Lightly-doped polysilicon bridge as a flow meter
Sensors and Actuators A
(1988)
T.S.J. Lammerink et al.
Micro-liquid flow sensor
Sensors and Actuators A
(1993)
N.T. Nguyen et al.
Low-cost silicon sensors for mass flow measurement of liquids and gases
Sensors and Actuators A
(1995)
H. Kuttner et al.
Microminiaturized thermistors arrays for temperature gradient, flow and perfusion measurements
Sensors and Actuators A
(1991)
B.W. van Oudheusden et al.
Integrated flow friction sensor
Sensors and Actuators A
(1988)

B.W. van Oudheusden

The thermal modelling of a flow sensor based on differential convective heat transfer

Sensors and Actuators A

(1991)

B.W. van Oudheusden

Silicon thermal flow sensors

Sensors and Actuators A

(1992)

J.W. Bosman et al.

Integrated smart two-dimension thermal flow sensor with Seebeck-voltage-to-frequency conversion

Sensors and Actuators A

(1992)

H.J. Verhoeven et al.

An integrated gas flow sensor with high sensitivity, low response time and pulse-rate output

Sensors and Actuators A

(1994)

K. Fricke

Mass-flow sensor with integrated electronics

Sensors and Actuators A

(1994)

D. Moser et al.

A high sensitivity CMOS gas flow sensor on a thin dielectric membrane

Sensors and Actuators A

(1993)

G. Stemme

A CMOS integrated silicon gas-flow sensor with pulse-modulated output

Sensors and Actuators A

(1988)

R. Kersjes et al.

An integrated sensor for invasive blood-velocity measurement

Sensors and Actuators A

(1993)

C.Q. Yang et al.

Monolithic flow sensor for measuring millilitre per minute liquid flow

Sensors and Actuators A

(1992)

D. Yu et al.

Microchannel pyroelectric anemometer

Sensors and Actuators A

(1993)

H.J.M. Geijselaers et al.

The dynamic mechanical characteristics of a resonating microbridge mass-flow sensor

Sensors and Actuators A

(1991)

F.R. Riedijk et al.

A dual-bit low-offset sigma delta analog-to-digital converter for integrated smart sensor

Sensors and Actuators A

(1993)

T.S.J. Lammerink et al.

Intelligent gas-mixture sensor

Sensors and Actuators A

(1995)

M. Esashi

Integrated microflow control systems

Sensors and Actuators A

(1990)

Sze, S. M., Semiconductor Sensors. John Wiley and Sons, New York,...

French, P. J., Gennissen, P. and Sarro, P. M., New micromachining techniques for microsystems. Proceeding of...

Gerlach, G. and Dötzel, W., Grundlagen der Mikrosystemtechnik. Hanser Verlag, München,...

G. Tschulena

Marktchancen sensoren

Sensormagazin

(1995)

Fuhrmann, Mikrosystemtechnik—ein wachsender Markt. Sensor Magazin, 1994, 4,...

A.F.P. van Putten et al.

Integrated silicon anemometer

Electronic Letters

(1974)

Cited by (364)

A highly stable, pressure-driven, flow control system based on Coriolis mass flow sensors for organs-on-chips
2024, Flow Measurement and Instrumentation
Stable delivery of liquids to microfluidic systems is essential for their reproducible functioning, especially when supplying flows to organs-on-chips – delicate living models that recreate human physiology on the microscale and thus can be used to reduce the need for animal testing. Most flow control systems are unable to sustain a robust and stable flow in longer experiments (>1 week), particularly those based on the ubiquitous syringe pump. Though easy to use, syringe pumps have no mechanism for actually measuring flow, let alone flow regulation with sensor feedback. We have developed a liquid delivery system based on the generation of flow by applying a constant air pressure to liquids in sealed containers. A flow of liquid is monitored by accurate measurement of mass flows (mg/min) using downstream Coriolis-based mass flow sensors. Measured mass flows provide fast feedback to integrated valves, with valves opening or closing slightly to increase or decrease solution flows to the organs-on-chips as required. This mass flow sensing principle is not affected by changes in the density, temperature, and viscosity of the liquids being displaced. This is in contrast to systems that use volumetric flow sensors, which require recalibration when these parameters change. The rationale behind using this principle for organs-on-chips, is that the stability provided by this flow control system allows for more control over growth of these mini-organs. We demonstrate the functionality of this system with three examples: 1) Fast stabilization (within seconds) under changing physical conditions; 2) Short-term stability (minutes to hours) of delivered flows in a microreactor with interconnected inlets; and 3) Long-term stability (>1 week) of cell medium flows to a living organ-on-a-chip. Two categories of organs-on-chips (OOCs) can be distinguished: 1) solid OOC are designed for three-dimensional cell or tissue constructs that interact with each other and their surroundings, and 2) barrier-type OOC contain a selective cellular barrier between two compartments as do many barriers in the body. The latter of these two types is the most challenging to culture and maintain as they are very sensitive to variations in flow and pressure surges. The flow control system presented in this work provides a great improvement compared to the use of syringe pumps and volumetric flow sensors in OOC studies. The novelty of this work lies in the long-term stability use of this system for organs-on-chips, maintaining stability for short to very long periods of time without compromising the barrier function of the organ-on-chip by pressure surges, bacterial contamination, or other undesired effects from the flow delivery system.
Recent developments toward microfluidic point-of-care diagnostic sensors for viral infections
2023, TrAC - Trends in Analytical Chemistry
This review article provides an overview of the application of microfluidics in point-of-care (POC) diagnostics, with a focus on rapid viral infection diagnosis. The introduction defines POC diagnostics and emphasizes the importance of rapid viral infection diagnosis. The history of microfluidics in diagnostics is discussed, along with a comparison to traditional diagnostic techniques. The principles of microfluidics and the basic components of microfluidic devices are presented. The advantages and disadvantages of microfluidics in diagnostics, including rapid diagnosis, low sample volume requirement, and integration with various detection technologies, are discussed. Examples of microfluidic diagnostic devices for viral infections, both commercial and research-based, are highlighted. Sensing technologies for viral infections are also explored, along with recent developments in this field. Moreover, materials and fabrication techniques used in microfluidic devices are described. Recent advancements in materials and fabrication techniques are also discussed. The challenges in microfluidic diagnostic sensor development are addressed too.
3D micro-structures for rarefied gas flow applications manufactured via two-photon-polymerization
2023, Vacuum
In this work, the two-photon-polymerization (TPP) fabrication technique was employed to manufacture three-dimensional (3D) micro-structures suitable for confined rarefied gas flows. Four different micro-tubes of simple geometry have been tested in vacuum conditions. The geometry of the micro-tubes was characterized via a focus-variation 3D optical measurement microscope, a digital microscope and an X-ray μ-tomography. Mass flow rate measurements of pressure-driven rarefied gas flows of nitrogen in slip and transition regimes were compared to theoretical values obtained via the kinetic theory of gases. The experimental measurements were realized via the constant volume technique. The good agreement obtained between experimental and numerical results proves the feasibility of using this fabrication technique to create well controlled 3D shaped microsystems for rarefied gas flows applications.
Fluid-compensated thermal flow sensor: A combination of the 3ω-method and constant temperature anemometry
2023, Sensors and Actuators A: Physical
Citation Excerpt :
Thermal flow sensors have a simple setup with only a few parts, and are suitable for mass production at affordable costs. But they are highly dependent on the thermal parameters of the fluid, e.g. thermal conductivity and heat capacity [3,4]. Fluid-independent thermal flow sensors being still not state-of-the-art offer advantages.
Thermal flow meters are widely spread in research, industry and daily life because of their simple handling and their cost efficiency. However, the signals of thermal flow meters are highly fluid-dependent. In fact, a conventional thermal flow meter can often not distinguish if the fluid or the flow has changed. A fluid-independent thermal flow meter is still a challenge in research. This article presents a new approach of a fluid-compensated thermal flow meter using a combination of the 3 $ω$ -method and constant temperature anemometry (CTA). The 3 $ω$ -method is used to characterize the fluid properties and CTA is used to characterize the flow. A key element of the proposed method is the combination of the theoretical models of the 3 $ω$ -method and CTA to obtain a fluid-compensated flow. This method has been characterized theoretically and experimentally, and it could be shown that it can determine the flow rate within $9 %$ , independent of the type of the fluid.
Non-invasive thermal energy flow rate sensor for turbulent pipe flows
2022, Flow Measurement and Instrumentation
Citation Excerpt :
Due to these limitations, research is necessary for new techniques – preferably simple, non-invasive, and low-cost systems [9]. Non-invasive thermal technology can be used for energy flow rate measurements, and has the advantages of easy installation, low-cost, a short response time, and low power consumption [28]. This technology is based on the heat transfer system that consist of two basic parts: a heater and temperature sensing elements.
A non-invasive thermal energy flow rate sensor based on a combination of transient heat flux and temperature measurements has been developed. A heat flux sensor and a thin-film thermocouple are covered by a thin-film electric heater and clamped onto the outer surface of a pipe. A one-dimensional transient thermal model is applied before and during activation of the external heater to provide estimates of the fluid heat transfer coefficient and thermal resistance. A parameter estimation code is used to estimate the unknown parameters by using the minimum root mean square error between the analytical and experimental sensor temperature values. The sensor was tested with two different fluids (water and diesel oil) and with two different pipe diameters ( $d = 25.4 m m$ and $d = 12.7 m m$ ). Experiments were completed over a range of Reynolds number from 2000 to 25,000 and a range of fluid temperature from 20 °C to 50 °C. Correlations for the fluid temperature and the fluid velocity were within ±0.13 K and ±5% of the direct measurements. The resulting non-invasive thermal energy flow rate sensor can be used to estimate a fluid velocity and a fluid temperature in heating and cooling pipe applications.
Automated and dynamic extrusion pressure adjustment based on real-time flow rate measurements for precise ink dispensing in 3D bioprinting
2022, Bioprinting
Extrusion-based printing relying on pneumatic dispensing systems is the most widely employed tool in bioprinting. However, standardized and reliable methods for process development, monitoring and control are still not established. Suitable printing parameters are often determined in a trial-and-error approach and neither process monitoring nor real-time adjustments of extrusion pressure to environmental and process-related changes are commonly employed. The present study evaluates an approach to introduce flow rate as a main process parameter to monitor and control extrusion-based bioprinting. An experimental setup was established by integrating a liquid flow meter between the cartridge and nozzle of a pneumatically driven bioprinter to measure the actual flow of dispensed ink in real-time. The measured flow rate was fed to a Python-based software tool implementing a proportional-integral-derivative (PID) feedback loop that automatically and dynamically adapted the extrusion pressure of the bioprinter to meet a specified target flow rate. The performance of the employed experimental setup was evaluated with three different model inks in three application examples. a) Continuous dispensing: Several runs of continuous dispensing showed that the PID-based pressure control was able to generate a steady flow rate more consistently and precisely than constant pressure settings. b) Adaptation to ink inhomogeneities: Deliberately created ink inhomogeneities were successfully compensated for by real-time pressure adjustments which profoundly enhanced the printing quality compared to printing without adaptive pressure. c) Process transfer to other nozzle types: Experiments with different nozzle types demonstrated the potential of the established setup to facilitate and accelerate process transfer and development. The present study provides an alternative approach for process design, monitoring and control by introducing flow rate as a main process parameter. We propose bioprinting processes to be based on flow rate specifications instead of constant pressure settings. This approach has the potential to save time by avoiding tedious parameter screenings and to introduce an active, real-time control over the printing process. Subjective influences by individual users during process development can be reduced and the process transfer between different devices and experimental setups can be facilitated and accelerated.

View all citing articles on Scopus

View full text