MEMS fluidic interface standards advance miniaturization of IVD technology and integration of specialized electronic devices.
|An integrated MEMS-microfluidics sensor device (actual size). Green printed circuit board layers support a silicon chip and acrylic microfluidic interface. A four-tube (850-micrometer Tygon) interface is attached to the right side.|
IVD technology is following a path trod by many other industries: as product sizes decrease, the level of integration increases. Integration allows the same machine or person to handle a smaller yet higher-level system using the same or smaller equipment. Miniaturization is a self-sustaining economic mode, because smaller integrated parts are more user friendly and easier to handle, store, and ship. A growing customer base enables producers to invest in a greater manufacturing infrastructure, which enables higher levels of automated assembly, which lowers the price per part while increasing quality. This creates a repeating virtuous cycle of improvement, distribution, and consumption of technology. For the purposes of this article, an industry exhibiting this virtuous cycle can be said to follow the miniaturization paradigm.
It is our observation that many parts of the IVD industry already follow the miniaturization paradigm, while others require further technical development to join the movement. In this article, lateral flow strips are used as an example of an IVD technology that is in the miniaturization paradigm; flow cytometers are an example of a technology that is not quite there. On the latter topic, we explain how the new SEMI standards for MEMS microfluidic integration could help get the cytometer onto a better miniaturization track. One specific need is the recruitment of microfluidic manufacturing equipment developers, and participation in interface standards activities by a broad cross section of industry members.
Decades ago, lateral flow strips had only one analyte, were very wide, required careful calibration, used stripes instead of microspots, and produced only qualitative, not quantitative, results. An industry developed over time that made great improvements in the quality of lateral flow strips while decreasing their size and cost. Roll-to-roll machines were developed to handle membrane rolls. The membrane thickness and uniformity was controlled. Automatic nano-spotters were developed. Plastic housings with standardized features and dimensions provided mechanical
|A thumb-driven syringe pump connects to flexible microtubing by means of a quick-connect luer fitting. The tubing connects to a MEMS biosensor device, part of which can be seen in the upper left corner of the image.|
support, protection, and humidity control. And inventive methods for integrating filters, sponges, reagents, and waste reservoirs into the total assembly were developed.
None of the improvements in lateral flow strip value and performance would have been possible without many interface developments. Each interface between a nitrocellulose membrane and filter or wick required experimentation with adhesives and production machinery. Interfaces between the outside world and the biochemically active micro world were developed. Improved interface designs combined with the integration of nearly all the difficult biochemical functions for assays has made it easy for customers to use lateral flow strip assays.
Dimensional design and interface standards can aid miniaturization if they are scalable. Such standards encourage investment in miniaturization because they reduce the risk of developing new machines for production and draw a broader array of companies into the miniaturizing industry.
In spite of great increases in value and integration within the lateral flow strip IVD marketplace, embedded electronics
|Figure 1. An illustration of the concept of pitch, diameter, spacing, and alignment holes. See Table I for a range of required values.|
in flow strips are still rare. One well-known challenge is the difficulty of producing thin layers of membrane in a non-roll format; another is the lack of machinery deigned to construct such electrically integrated devices. Microfluidic IVD devices, while newer to the marketplace, have a greater likelihood of integrating electronic sensors, actuators, and controllers.
Many kinds of microelectronic devices are used in diagnostic systems. Digital computational power is employed to enhance clinicians’ ability to make sense of complex patient data. Computers also help direct the activity of machines that process samples and measure biochemical and physical properties of samples. But beyond basic computing
|A MEMS biosensor assembly. The green ribbon is a flexible printed circuit board. Four flexible microtubes extend downward from the center, where the biosensor electronic chip, microfluidic cover, and four-tube adapter are located.|
capability, the systems have smarts and dexterity that allow varying degrees of autonomous sample processing and analysis. The smarts come from electronic sensors and data processing; dexterity comes from electronic actuators. Electronic actuators, themselves, require their own sensors to verify that they are performing the actuation function. An example of a smart actuator is an electronically controlled valve with sensors to monitor fluid flow rate through the valve.
The majority of smart IVD systems are still too large and expensive to be practical or usable by general consumers. However, there is progress. Autonomous diagnostic systems are now benchtop size, and getting smaller. Once they are comparable in size and cost to a disposable lateral flow strip test, autonomous diagnostic systems will be ready to enter the mainstream consumer market. There is no intrinsic impediment to this outcome, aside from time, money, and technology development.
To enhance and encourage the existing miniaturization paradigm in IVDs, a new microfluidics interface standard, SEMI MS9-0611 has been published.1 This standard describes interface designs and material selection, provides a set of dimensional requirements needed for components to fit together, specifies geometries, and enables tooling and
|Figure 2. The solid arrows show the flow path of cells and fluid through an electronic integrated cytometer. The dotted double arrows indicate electrical signals passing to and from subsections of the system. The dashed lines suggest additional functionality that would be possible with microscale high-density fluidic interfaces.|
service providers to develop cost-effective handling and assembly systems. The intent is to support growth and improvement of the manufacturing base that will build smaller electronic fluidic systems in shorter design cycles for lower cost.
SEMI MS9-0611 applies to a relatively narrow niche of high-density permanent connections between microfluidic devices. Even within this narrow scope, the standard can use significant improvement and modification, and incorporation of better manufacturing and quality measures.
The specification part of SEMI MS9 has two types of requirements, the first of which is more flexible and less prescriptive, the second of which requires the manufacturer to select interface dimensions from a short list of possible values.
|An integrated MEMS-microfluidics biosensor device (scale: ~20:1). The blue-green rectangular sensing areas can be seen through two layers of acrylic microfluidic structures and some adhesive. The silicon-based biosensor chip is electrically connected to the green PCB. A four-tube interface is attached to the right side.|
Manufacturers who wish to comply with Part A of the standard must publish dimensional and material information about their products. Also, they must design their products to be generally compatible with interface specifications. This is a minimal set of detailed requirements for specifications that a manufacturer must include on the product data sheet to describe the product’s interface with the outside world. The specs include port size (inner diameter or ID), port spacing, port location, number of ports in a row or array, any physical alignment features, and the material composition of the wetted flow path.
Manufacturers who wish to comply with Part B must select from sets of specific port spacings and dimensions. A typical interface could have eight parallel fluidic tubes with a center-to-center spacing of 0.500 mm and an ID of 0.250 mm. A four-port example is shown in Figure 1.
The red rectangular shape in the middle is the top surface of a sealing plane between two permanently connected devices. Four circular ports are shown in the red surface, each carrying a different fluid inlet or outlet. The large green rectangle is a circuit board upon which the electrofluidic device is placed. The grayish region surrounding the red center is a keep-out volume, whose top surface is coplanar with the red sealing surface. The two alignment holes at either end of the circuit board allow optical- or pin-based alignment using automated assembly equipment.
Other nondimensional requirements include specification of:
An emphasis is placed on ensuring that the entire architecture’s dimensions are scalable downward.
Microfluidics standards are needed for manufacturers of components. The need is driven by technical requirements, re-engineering, reliability, assembly, and testing, each of which can significantly impact cost, development time, reproducibility, compatibility, and quality. For some critical applications, it is important to specify materials and surface treatments that do not interact with the fluid. We can also look at the connection of tubing of various sizes as a challenge when bidirectional flow is required and unswept volumes are undesirable, or where entrapment of air may be
|An acrylic microfluidic assembly for integration of MEMS with microfluidics biosensor devices (scale: ~20:1). At the center is a four-port manifold to which four Tygon tubes measuring 850 micrometers in diameter are bonded.|
a concern. Solutions are known yet may not be applied across the microfluidics industry. A small leak or air bubble can lead to blood clotting when medical instruments are developed. Such effects can be catastrophic, and tolerancing in the design and manufacture of fittings and interfaces is important to minimize such defects. Identification of critical manufacturing parameters and their control limits are initially required. It is important to perform instrument calibration in manufacturing as well as to maintain good process control of key parameters. Finally, the type of connections and splicing of tubing or interfaces should be designed to meet the needs of the application.
Generally, standards need to provide workable definitions of interface dimensions and attachment methods so that a plurality of manufacturers can provide unique technology with minimal engineering effort. These interfaces need to have enough definition that a manufacturer can, in fact, use them, but enough flexibility to allow a variety of applications to fit into the overall scheme. The interface design should be scalable across a range of fluidic channel dimensions and parallel channels. Another step is to provide manufacturing and assembly service providers with enough information that they can develop automated assembly tooling and processes. Furthermore, the designs should use materials and processes that do not damage on-board biochemical molecules and take advantage of the properties of plastic for molding and bonding.
A broadly used standard in IVD applications is the Well Positions for Microplates ANSI/SBS 4-2004.2 This standard enables the design of automated machines based on the regular location of a two-dimensional array of wells on a
|Figure 3. A cross section diagram of a typical temporary seal with edge clamps.|
surface measuring approximately 8 x 13 mm. This standard has enabled a systematic increase in the density of wells, with a total count increase from six to 3456. However, the total area of the plates has not changed, and the microplate standard does not provide for integrating any electronic devices into the wells. Combining other functions such as valves, pumps, and sensors has proceeded with no generally accepted industry format.
Some researchers have tried to provide an integration format that could be accepted in industry, as part of the thrust to develop lab-on-chip technology. An early example published in 1998 uses the printed circuit board (PCB) construction standard, as well as standard spacing between package electrical leads, as a basis for standardization of microfluidic device integration.3 This concept provides all of the needed elements for integration of fluidics with electronics. However, it is based on an old through-hole PCB design format and needs updating. PCB technology has rapidly shrinking design rules because it is in the miniaturization paradigm. As through-hole technology gives way to surface-mount, chip-on-board, and flip-chip packaging methods, the amount of space needed on PCBs continues to shrink. Since the electronic packaging industry is rapidly improving its areal density, fluidic interface standards also must increase their areal density to achieve commercially relevant interface dimensions. A very recent lab-on-chip standards review article provides an up-to-date consideration of this area from the fluidics standpoint.4
This increasing areal density is not really a problem if the standards community can embrace the need for scalable designs and rules. That is, microfluidic interconnection concepts that apply at-length scales ranging from 10,000 micrometers to 1.0 nanometers and below. Table I summarizes the technology readiness of microfluidic standards over all scale regimes.
|Table I. The first column divides the miniaturization paradigm into three broad categories: meso, micro, and nano. The other columns show trends that parameters such as cost and integration follow. The last column shows that relatively few standards are in use at the nanoscale compared with the mesoscale.
The challenges of interfacing electronics with fluidics are described in broad terms in the “Guide for Design and Materials for Interfacing Microfluidics Systems,” (MS6 2008).5 While a guide does not have the quantitative precision of a specification, it does contain nomenclature and interface descriptions that form the basis for further work. This guide covers a broad range of possible materials and applications, including ultra high purity, high pressure, vacuum systems, ceramics and steel fixturing. Potential IVD applications are a small fraction of the total application space.
Integration is going on in the electronics and fluidics layers, and in devices that access both of these. SEMI MEMS standard MS7 20076 describes these layers in some detail, systematically giving them the names of fluidic routing card, circuit routing card, and MEMS electrofluidic integrated circuit, respectively. It also provides detailed designs including one for a 64-channel fluidic MEMS Electro-Fluidic Integrated Circuit (EFIC) and mating fluidic routing card device. The nomenclature proved to be helpful, but that standard has not yet been validated. Several issues need resolution, including
Integrating high-performance biodetection systems into tinier and less expensive form factors requires technical advances in fluidic interface design and standardization. A representative example of such an IVD system is a flowing cell counter, or cytometer. See Figure 2 for a block diagram illustrating the main functions it performs. The functional blocks are connected by fluid lines and electrical lines. Each fluidic, electrical, and data interface can benefit from increased standardization.
Currently, cytometers contain hardware for generating and measuring flow and pressures; sensors for counting cells; actuators for deflecting counted cells; and sample inlet and outlet handling for incoming liquids and sorted outgoing material. Though the active counting volume of these systems is approximately 1 mm3, the system total volume is 106. One reason the space is not efficiently used is the many fluidic connectors needed between the multiple types of hardware. Another reason is that each fluidic connector may carry only a single tube or flow path. Yet another is that the electrical components in the system are not well integrated but, rather, have
|Figure 4. Two types of permanent microfluidic tubing interconnections are shown: insertion on the left and butt joint on the right.|
collections of wires and comparatively bulky connectors extending out from the system to be plugged into a master circuit board.
Each arrow in Figure 2 represents a connection—fluidic, electrical, or data—needed for the system to function. Integration of the data and electronic components is readily available from industrial suppliers. Integration of the fluidic connections, however, is a relatively new technology and is not performed at the same scale.
Many research groups are actively working on integrating cytometers. Two have recently demonstrated cytometers with electronic detectors inside the microfluidic flow channel. One example uses a CMOS photonic detector;7 another, a magnetoresistive detector.8
Clearly, advanced technical work is taking place in lab-on-chip technology. Adapting this technology to industry requires additional planning and resources, including standards. The existing standards described above provide a good foundation; however, they are somewhat broad for advancing the specific needs of IVD miniaturization and the lab-on-chip industry.
The SEMI-MS9 standard helps IVD applications by focusing on a few materials and processes that are likely to be popular with industry. Moderate operating and burst pressures (neither ultra-high vacuum nor high absolute pressure) are encouraged. Molded plastic is called out as a desirable material because of its low cost, biocompatibility, and ease
|Table II. Sampling of companies with developments or products that enable or enhance the IVD miniaturization paradigm. The list is incomplete and is meant only to show the existence of diverse company types and locations.
of fabrication. Permanent connections (glue or melting) are favored over reusable methods because they reduce product size, cost, and complexity. Together, these materials and processes support production of consumable diagnostic technology. Consumable diagnostic products help reduce health risks related to re-use and sterilization of healthcare equipment.
There are two main kinds of information in “MS9, Design Guides and Specifications.” The specifications of MS9 were discussed above. Specifications are extremely valuable if the relevant industry adopts them. Because they are highly specific and legalistic, however, they must have a limited scope of use. Design guides, by contrast, provide help without requiring anyone to do any particular thing with a design. Consequently, design guides are consulted more frequently.
A typical macroscale connector (threaded pipe fitting to a manifold, for example) serves three functions in one:
The macroscale connector’s size allows it to perform all these functions without the need for extra features. However, as the bore dimensions and pitch of the fluidic connectors become smaller, these functionalities have to be decoupled. The small dimensions do not provide sufficient mechanical strength or space for mechanical alignment features. The challenge lies in decoupling these functions using additional features while simultaneously maintaining the smallest possible physical dimensions of the connector.
To create a connector or interface that provides leak-free coupling for an array of holes, bore diameter, pitch, and method of attachment are critical factors. Temporary seals are typically harder to achieve than permanent seals, as temporary seals involve some form of gasket and mechanical clamping mechanism (Figure 3). It is difficult to apply a uniform sealing force over the gasket area, which makes the setup prone to leaking. Permanent seals are more desirable and are attained using adhesives or other bonding methods providing stronger, leak-proof interconnections for smaller bore diameters and pitch.
|Figure 5. Cross-section of fluidic routing card above, PCB below, and electrically active sensor or actuator in between.|
The challenges in making a clamped interface can be overcome by using a permanent tube interface. Figure 4 shows examples of how this can be achieved.
When integrating fluidics with electronics, typically a two-step approach is necessary. The first step delivers fluid from a conventional storage system (bottles, vials, or other reservoirs) to the device; the subsequent step involves a coupling to deliver fluid to a silicon sensor surface.
While the first step may be approached with tube-connector arrangements (Figure 4), the second step is best achieved using laminar connectors in the form of routing cards (Figure 5). This geometry is coherent with electronic printed circuit boards and allows intricate fluid delivery networks to be formed in confined spaces. Multilaminated routing cards can be formed in glass or plastic with any number of layers. Commercial sources do exist for custom and off-the-shelf designs with published design rules. One such vendor, Aline, published a description of its offering.9 Routing cards may interface with packaged silicon die as shown in Figure 5. A multifunctional type design is shown in Figure 6.
A commercial standard is only validated through adoption by the marketplace and industry, which happens over five- to 10-year periods. In the mean time, there are many opportunities to participate in and influence the direction of standards for IVD integration. Potential standards participants will ask the question: will this standard be used? If so, when? What commercial forces will drive and justify the effort needed to invigorate the miniaturization paradigm for IVDs?
There is already significant commercial activity, some of which is in development, as well as real high-quantity production. One of the companies authoring this paper produces parts and subcomponents designed to comply with the MS9 standard. Many other companies participated in generating the standard and certainly have capabilities to produce parts with only small design modifications. Commercial organizations that may participate in or benefit from developing the IVD miniaturization paradigm through standards activities are listed in Table II.
At least three major manufacturing functions are lacking in the IVD miniaturized high-density interconnect industry:
|Figure 6. Cross section of multilayer fluidic routing card having permanent fluidic connections between multiple microscale sensors, valves, and pumps.|
pick-and-place machines designed for IVD microfluidic components and providers of automated assembly services and automated test equipment. The existence of IVD component interface standards will encourage equipment manufacturers to increase their investment in the integrated microfluidics marketplace. Some additional standards or modifications also may be needed. Each of these areas provides potentially large market leadership opportunities for companies that can supply these needed machines and services. But will the market be attractive enough to overcome the many risks involved?
One eye-catching opportunity announced in early 2012 is the Tricorder X Prize. “The Qualcomm Tricorder X Prize is a $10-million competition to stimulate innovation and integration of precision diagnostic technologies, making definitive health assessment available directly to health consumers,” notes the website. “These technologies on a consumer’s mobile device will be presented in an appealing, engaging way that brings a desire to be incorporated into daily life.”
Achieving this goal will certainly require highly integrated IVD technology. Miniaturization standards may enable this and related progress.
1. "Specification for High Density Permanent Connections between Microfluidic Devices," SEMI MS9 (2011).
2. Well Positions for Microplates, ANSI/SBS 40 (2004) – ANSI/Society for Biomolecular Screening.
3. P. Galambos, and G. Benavides, “Electrical and Fluidic Packaging of Surface Micromachined Electro-Micro-Fluidic Devices,” in Micro Total Analysis Systems, ed. J.D. Harrison and A. van den Berg (Banff, AB, Canada, 1998).
4. H. van Heeren, “Standards for connecting microfluidic devices,” Lab on a Chip 2012 12: 1022.
5. "Guide for Design and Materials for Interfacing Microfluidic Systems," SEMI MS6-0308 (2006).
6. "Specification for Microfluidic Interfaces to Electronic Device Packages," SEMI MS07, (2008).
7. Y. Hosseini and K. Kaler, “Integrated CMOS optical sensor for cell detection and analysis,” Sensors and Actuators A: Physical 157 (2010): 8.
8. J. Loureiro et al., “Magnetoresistive chip cytometer,” Lab on a Chip, 2011, 11: 2255.
9. L. Levine, “Developing Diagnostic Products: Polymer Laminate Technology,” MD+DI, 31, 2 (2009).
Mark Tondra is president and founder of Diagnostic Biosensors. He is also co-chair of the SEMI MEMS – Microfluidics interface task force. He can be reached at firstname.lastname@example.org.
Mark Crockett is president and founder of MEMSmart. He is also co-chair of the SEMI MEMS – Microfluidics interface task force. He can be reached at email@example.com.
Proyag Datta is a senior microfabrication engineer at Second Sight Medical Products. He can be reached at firstname.lastname@example.org.