• Projects/Funding
  This page contains data on projects enabled by SCIMPS and associated funding

• Research Projects
  The following projects are research projects led by each participant of the consortium

• Institute for System Level Integration : Design and modeling of microscanner.
• Heriot Watt University: High bumps for micro-LED flip-chip array packaging.
• Heriot Watt University: Electrostatically induced patterning.
• Heriot Watt University: Hermeticity methods for microsystems.
• Heriot Watt University: Health and Usagee monitoring Microsystems (HUMMS) for failure detection and health management for aircraft wiring.
• University of Glasgow: Integrated array system.
• University of Strathclyde: Centre for Microsystems and Photonics: Integrating integrated optics and microfluidics for sensing.

Institute for System Level Integration: Design and modeling of microscanners.

An early focus of the research work to be carried out by iSLI, involves developing micro-scanning devices for use within miniaturised dynamic optical systems being developed by the other member of the consortium. This has led to the development of a number of variants of mirror scanners, together with additional prototypes designs.

Working with the Micro-Scale Sensors Group within the University of the West of Scotland [www.uws.ac.uk], the Institute has proposed a simple structure for an optical microscanner, composed of two piezoelectric bimorph beams that are used to actuate a mirror plate. We have proposed a complete analytical model of the piezoelectric device that lead to the profile of the structure and the tilt of the mirror. This model can be can be used very rapidly to study the effects of geometrical parameters and material properties on the performance of the micro-scanner, without having to revert to the use of computationally expensive finite element modelling at the outset. This understanding forms the basis of further behavioural modelling and VHDL integration, allowing rapid system level simulation and optimisation of these structures. This work has been the subject of papers published in IEEE’s 2007 Symposium on Design, Test, Integration and Packaging (DTIP’07) and in the Journal of Microsystem Technologies.

FEM result of a piezoelectric micro-scanner.

The iSLI MEMS design team have developed a new MEMS scanner fabricated by a single layer SOI process; both micro mirror and vertical actuator are fabricated in the same layer. The scanner is based on thermomechanical actuation, with a thermal response time of 50 ms. Tests show a maximum static tilting angle of 5 degrees, achieved at 18 volts and 23 mA; corresponding to a 10 degree optical scan angle. The scanner can be operated from DC up to its mechanical resonant frequency of 2.19 kHz; at which frequency a mechanical tilt angle of 8 degrees has been demonstrated. This lower frequency operation meets the requirements of optical coherence tomography (OCT) systems, amongst other applications. Operation at the resonant frequency, with 16 degrees of optical scanning, makes it useful for application in a high speed confocal microscope. Theoretical analyses of the static and dynamic performances of the scanner have been conducted and a good agreement with experimental results was reached. Scanners of this type have potential applications in many fields including endoscopic imaging systems, laser based displays and miniature bar code scanners.

ISLI thermally actuated mems scanner

Heriot Watt University: High bumps for micro-LED flip chip array packaging.

Using a Micro-LED array as a micro-light emitter has many potential applications in many areas such as microscopy, lab-on-chip screening and mask-free lithography etc. The Institute of Photonics (IoP) has developed a 64x64 micro LED array which could be used for such purposes, as shown in Figure 1.The challenge when increasing the size of the array whilst maintaining a compact component footprint, is that the necessary interconnects to the drive the electronic circuitry must become smaller and have a smaller pitch. Traditional methods, such as wire bonding, become very complex, laborious and unreliable. Therefore, the preferred method of interconnection is Flip-Chip Bonding (FCB).

FCB makes it possible to integrate the micro-LEDs in a high density matrix-addressed format so that each LED can be operated independently. FCB can also improve the light output power and thermal capabilities of the overall component. However, commercially available gold wire bonders cannot produce Au bumps with bump diameter of a few tens of microns whilst maintaining the fine inter-bump spacing of 12µm. In addition, commercial suppliers have failed to produce a perfect yield using wire bonder bumps and ultrasonic FCB.

The initial research investigated the feasibility of an adapted UV-LIGA process. Au bumps 5-10µm in height, 30µm in diameter with a 12µm pitch were fabricated in a 16x16 array as below. A 3D plot obtained from the NewView 5200 white light interferometer displays the height uniformity of the electrodeposited bumps. Bump height uniformity is important when requiring a high interconnection yield when FCB.

FCB tests have been performed using thermo-compression with these prototype Au arrays on gold plated glass samples. On visual inspection the bonding has been successful. Shear and pull strength of the bumps still need to be carried out.

Presently the issue remains the compatibility of this fabrication with the IoP fabrication process for the micro-lens array. A bump transfer method and electroless Au fabrication are two options currently under investigation.

Heriot Watt University: Electrostatically induced patterning.

We have developed a technology for the microfabrication of patterns in planar and non-planar (3D) surfaces in polymer using electrostatic pressure. The developed technology is protected by a GB patent application #0724718.2 and potential partners are now welcome for cooperation on the further development and commercialization of the technology.

There are several advantages of electrostatic induced method over traditional photolithography technology. Firstly, there is no expensive optical equipment involved since no photolithographic equipment is needed. Secondly, any polymer, in principle, should be suitable for this technique. It has been possible to pattern polymers beyond the thickness specified by manufactutrers. The fabrication cost can thereby be greatly reduced since thick photosensitive resists can be quite expensive. Thirdly, there is no development process needed due to the one-step pattern formation after annealing of polymer.

We have recently successfully extended this technology into the microfabrication area and made patterns even on 3-D surfaces which are impossible for traditional photolithography. The potential applications of this technology could be the follow areas:

A brief overview and some results of the developed technology are summarized as follows.

1. Operating principle and experimental set up
Electrostatically induced patterning is a non-traditional lithography method to make patterns in polymers. The method can be used from the meso-scale to the nanoscale. The electrostatic pressure redistributes the polymers on substrate in an ordered or self-assembled form. Fig.1 shows the normal setup used to perform the electrostatically induced lithographic process.

2. Microchannels fabricated in polymer
Fig.2 shows the deep microchannels with a depth of up to 83µm that have been achieved in a novolak based polymer deposited with an initial thickness of only 25µm. The whole process for this fabrication took about 1 hour also the process has not been optimised timewise. Fig.3 shows the microchannels with a depth of 33µm in a UV curable adhesive. The process time for this fabrication is very fast and the patterning could be achieved in 10 minutes.   

Fig.2. SEM image of the fabricated microchannels with a width of 100µm and depth of 83µm (left), 2-D and 3-D surface profile of the microchannels measured by Zygo optical interferometer (right). 

Fig.3. SEM images of the microchannels with a depth of 33µm fabricated in a UV-curable adhesive.

Heriot Watt University: Hermeticity test methods for microsystems .

There are several problems associated with the traditional test methods for hermeticity when applied to microsystems. The helium fine leak test is often used in conjunction with a gross leak test, such as the bubble test, to determine the hermeticity of packages. When applied to microsystems with cavity volumes of less than 50mm3, the minimum leak rate detectable by the helium fine leak test becomes insufficient. During fine leak testing packages are pressurised or bombed with a specified pressure of helium for a time defined by the military standard MIL-STD-883. The package is then transferred to a mass spectrometer where the helium leaking out of the bombed package is detected and a leak rate calculated.  This test is conducted prior to the destructive gross leak test. The lowest volume stated in the military standard is for cavities less than 0.05cm3, two orders of magnitude larger than a typical MEMS cavity of 0.1 mm3. As cavity volumes decrease the maximum detectable leak rate of the fine leak test becomes less than the minimum leak rate detectable by the gross leak test, leaving a gap in the detectable range.

Figure 1: Leak rates detectable from micro-cavities.

The minimum leak rate using the helium test can be lowered to 10-11atm.cm3.s-1 if the cavity is sealed in a helium environment. This is however still 7 orders of magnitude larger than the theoretical minimum leak rate of a typical microsystem. A package with a cavity volume of 100nl is subject to leak rates as low as 6e-18 atm.cm3.s-1 over a five year period.

Several other test methods are being used currently to test hermeticity. The Q-factor test is commonly used for testing MEMS resonant structures as this test method requires no further test structure and no additional methods to establish the hermeticity of the seal. The resonance peak is higher and sharper the lower the pressure, meaning that the Q-factor of the device directly indicates the ambient pressure. A leak rate of the device can therefore be determined non-destructively by monitoring the change in Q-factor over time. The test can however only be applied to MEMS containing free standing structures. It can also be difficult to calibrate, insensitive to small pressure variations and suffers from sensor drift due to material instabilities.

The deformation of packages method can be used to test the hermeticity of microsystems with suitably flexible membranes non-destructively. Membranes will deflect when the pressure inside the cavity differs to that outside. If the membrane begins to equalise over time it is clear that the package is not hermetic. Packages can be calibrated and deflections measured using various optical means. This technique currently suffers from poor sensitivity, reliant on the resolution of the optical equipment and requires extensive calibration for each device type and membrane thickness.

Figure 2: Deformation of Packages
De Wolf I, J.A., De Moor P, Tilmans HAC, Marchand L, Hermeticity Testing and Failure analysis of MEMS Packages. IPFA 2007. 14th International Symposium on the Physical and Failure Analysis of Integrated Circuits, 2007: p. 147-154.

Copper test patterns have also been used to determine the hermeticity of MEMS cavities. This involves measuring the change in the optical transmission of copper deposited within a cavity. Copper is opaque to near infrared wavelengths but as it oxidises it becomes transparent. This means that the oxygen partial pressure can be determined by non-destructive optical means and the associated leak rate of the package determined. In studies by Geuissaz this method has been reported to measure leak rates down to 5e-16 atm.cm3.s-1. This technique is being further developed by the Institute of Photonics at Strathclyde University and MISEC at Heriot-Watt University.

Figure 3: Copper Test Pattern
 F.Gueissaz, Ultra low leak detection method for MEMS devices, 18th IEEE International Conference on MEMS 2005: Miami Beach, Florida, USA

MISEC is currently developing a method using test structures to monitor the change in resistance associated with a change in ambient pressure. When a constant current it supplied to a resistive structure the material will become hotter leading to a change in the temperature coefficient of resistance, TCR, and thermal conductance of the material.

Figure 4: Temperature dependence of thermal conductance on aluminium Oxide.

A free standing structure will be affected only by convection to the surrounding ambient as heat conduction due to radiation is negligible and heat conduction to the surrounding materials can be kept to a minimum as a result of structure geometry. By monitoring the TCR or thermal conductance of a free standing structure, the pressure surrounding the structure can be determined and the leak rate of the package calculated.

The development of the FTIR hermeticity test method for microsystems is also taking place at MISEC. A tracer gas, usually nitrous oxide as it has high peak absorption at infra-red wavelengths and a similar molecular mass to air, is used to bomb a package before FTIR spectroscopy is used to determine the concentration of gas in the cavity over time. This test is limited to microsystems that have a cap such as thinned silicon that is transparent to IR radiation. MISEC is currently working to test this technique on smaller packages and to eliminate problems associated with internal reflections from different surfaces.  

 Figure 5: Principle of Hermeticity Assessment by FTIR Spectroscopy.
Veyrie D, Roux JL, Pressecq F, Tetelin A, Pellet C, A new method to assess the hermeticity of MEMS micro-packages, Proceedings of the 5th ESA round table on micro/nano technologies for space, 2005, Noorwijk, The Netherlands.

Heriot Watt University: Health and Usage Monitoring Microsystems (HUMMS) for failure detection and health management for aircraft wiring.

At present there is growing industry focus in the design, fabrication and implementation of health and usage monitoring systems in environments where routine maintenance is difficult perform. This is due to difficulties in accessing specific areas, which may be too remote, difficult/dangerous to access and cost prohibitive.

Advances in recent years of MEMS design and manufacturing techniques mean that it is more realistic to create and integrate micro-sensors possessing different functionalities on a common platform. This leads to the possibility of in- situ testing or monitoring of aircraft wiring and interconnect through the design and integration of micro-sensors (monolithically or by hybrid combination). These micro-sensors (also known as Health and Usage Monitoring MicroSystems) will occupy a small volume, enabling the insertion of these systems into key areas of the aircrafts electrical wiring and interconnect systems where high rates of aging and failures occur.

 
Figures 1a and 1b: Aircraft avionics engineers using specialised automated test equipment to check and locate wire and interconnect faults.

One advantage of using HUMMS is that monitoring of the wire and interconnect can be performed whilst the aircraft is in flight rather than just performing maintenance checks when the aircraft is grounded. It is widely reported that there are difficulties in detecting intermittent effects when the aircraft is checked when grounded. The reason for this is that intermittent faults present themselves when the aircraft is in flight and subject to harsh environments such as temperature, vibration, stress and humidity. Automated test equipment can detect opens and shorts in the wiring, but this cannot be used when the aircraft is in flight.

One solution being investigated by MISEC is the design, manufacture and implementation of a wide bandwidth current sensor. Such as sensor is based on macro scale rogowski coils or current transformers. The sensor itself works on the principle of magnetic coupling between the wire under test (WUT) and the sensor itself.  An alternating current in the WUT induces a voltage in the sensor that is proportional to the time rate of change of current. To obtain a voltage signal proportional to the current waveform, integration of the induced sensor voltage is required. This can be achieved by choosing the load resistance that appears across the sensor such that the ration of the sensors inductance to the load resistance is greater than the period of the waveform being measured. This eliminates the need for using additional op-amp integration. Advanced signal processing methods such as Wavelet Transforms are being considered in combination with the rogowski coil as they can detect arc fault waveforms in the presence of noise and other sources of electromagnetic interference (EMI).

Other advantages of the rogowski coil include:

• The measurement of fast transient currents of nano-second rise-time, that is typical of arc faults commonly seen in aircraft wire and high voltage cable insulation.
• Good linearity due to the lack of magnetic material.
• Galvanic isolation between WUT and the sensor.
• The ability to use Microsystems manufacturing techniques to shrink the sensor in size so that it can be located inside wire interconnect or other small and hard to access areas
• Lower measurement errors due to increased manufacturing accuracy.
• Cheaper to mass manufacture.

MISEC has known expertise in the design, micro-manufacture and test of micro-inductors in various design topologies through the use of UV-LIGA. These capabilities have been utilised to create a series of planar rogowski coil sensors. 3D torroidal style sensors are planned for manufacture in the near future. This will allow these sensors to be placed in certain locations where there is a known failure issue.

Figure 2a and 2b: Original concepts of how rogowski coil and additional HUMMS sensors could be integrated and positioned within aircraft wire and interconnect. 2a reflects the idea of a Smart Plaster rogowski sensor and environmental monitoring sensors that can be wrapped around the cable insulation. 2b shows the idea of a smart disk that could be placed inside the wire interconnect.

Additional micro-sensors such as humidity, temperature and strain are being designed and manufactured through ISLI and BUTE through the PATENT DfMM Network of Excellence. Such sensors are critical to the determination of the accelerated aging and wear and tear experienced by the wire insulation and interconnect. Elevated humidity levels on aircraft lead to accelerated corrosion rates of the metallic interconnect (causing higher levels of intermittent faults), whilst the combination of humidity and temperature lead to higher degradation to the chemical, mechanical and electrical integrity of the insulation. Strain gauges will enable to detect areas of wiring that are placed under undue stress. These sensors will enable the prediction of potential failures and enable corrective maintenance to be carried out, and provide massive cost savings to the aviation industry.

Figure 3: Planar rogowski coil sensor manufactured using UV-LIGA.

Figure 4: SEM of the Planar Rogowski Coil fabricated using the MUMPS Process.

University of Glasgow: Integrated sensor array system.

Microfabrication has seen continuous improvement in the last 40 years allowing integrated circuits to be faster, more accurate and reliable.  In addition, microfabrication technologies have been widely adopted in other spheres to create micro-electromechanical systems (MEMS), laboratory on a chip (LOAC) and many types of sensor.  This has often entailed adding new materials or using devices and technologies that are more usually intended for electronics applications in unexpected ways.  As a consequence, microelectronics technologies are now widely used in sensor applications as diverse as cameras, biomedical implants, microbiological sensing and environmental monitoring.

We have created an integrated sensor array system that has sensors built from photodiode structures. This system is based on the principle of the light-addressable potentiometric sensor (LAPS). LAPS as it is now commonly known uses a photodiode to monitor changes in surface charges on a chemically sensitive surface. In our design, we are using the top passivation layer of an integrated circuit (silicon nitride) as the chemically sensitive surface. Silicon nitride is chemically sensitive to hydrogen ions, hence it allows us to measure pH. We intend to use this device to monitor the changes in extracellular pH of live tissue culture. Changes in extracellular pH is an indication of the state of cell metabolism. This is an important indicator as unusual quality or rate of metabolism are often linked to early onsets of diseases and other potential detrimental effects to cell viability.

University of Strathclyde (Centre for microsystems and photonics): Integrating integrated optics and microfluidics sensing.

The broad aim of this work is to develop miniaturised, advanced systems for the measurement, delivery and in-situ analysis of biological analytes, which will combine optical measurements with lab-on-a-chip style microfluidics.
Evanescent field sensors are the most common way to combine waveguides with microfluidics, however typically less than 10% of the guided light, is guided in an evanescent mode. The work in this project takes a different approach, with the aim of developing a system in which microfluidic channels intersect with the waveguides at about 90 degrees, in order to maximise the light-analyte interaction. Multimode guides with a width of 50um are used to maximise the light in the waveguide.
Both waveguides and microfluidic channels are formed in an acrylate based polymer. The polymer is formed when the monomer is exposed to UV light, which causes it to cross-link. The features can therefore be defined using photolithography. In this work a three layer system is used consisting of a cladding, a core layer and a final cladding layer. The index of the core layer is approximately 1.52. Rib waveguides are patterned into the core layer with gaps to coincide with the microfluidic channels, which are defined in the top cladding layer, so that they intersect with the waveguides at 90 degrees. Figure 1 shows a microscope image of the microfluidic channel. The waveguides can also be seen underneath the cladding layer.
A diagram showing the optical set-up is shown in figure 2. Light from either a He-Ne or an Argon Ion laser is coupled into a multi-mode fibre and launched into the waveguides. The output of the waveguides is focussed using a microscope objective, or collected directly by a fibre and measured using a CCD spectrograph. A PDMS over-layer is used to cover the microfluidic channel, and to provide an inlet and outlet for the microfluidics. This will be connected to a syringe pump for accurate control of the flow rate.