A. Nanofabrication Techniques
Initially progress was slow due to the challenges of using innovative, nanofabrication technology being applied to difficult materials, requiring significant development by the University of New South Wales (UNSW), Australian National Fabrication Facility (ANFF).
A number of concepts were investigated commencing with a simple nano-metre based equivalent of the HSE/NPL Mark II Slide. At the time of development, it was not possible to have different groove depths on the same substrate, so the approach was changed to single ‘chips’, each with their own individual groove depths.
Around 20 different substrates and 5 different coatings were investigated so that the respective refractive indices would lead to practical groove or ridge heights.
Various different configurations were investigated to successfully integrate these materials.
One configuration was chosen from a total of five different configurations where phase objects were formed by nanofabrication processes including photolithography and Reactive Ion Etching which were developed as a result of extensive testing. This required significant modification and fine tuning of multiple processing steps to provide stable, accurate, precise and uniform chips.
The individual processing steps to create the quartz chips are as follows:
Step 1
- Photolithography of alignment marks using Photo Lithographic Mask #1 and a negative resist, so as to form alignment marks used to align a series of radial and circular guidelines with the phase objects that are not visible.
- Titanium (Ti) & aluminium (Al) deposition and lift off, leaving alignment marks.
Step 2
- Photolithography of the phase object targets using Photo Lithographic Mask #2 and a positive resist.
- Reactive Ion Etcher (RIE) etch of phase objects, which are the 20 parallel lines, each 1µm wide and 100µm long.
Step 3
- Measurement of groove depths is critical during fabrication and for later choice for use in the Test Slides. Two techniques are used.
- Dektak measurements to determine the etch depths, of the 102 mm diameter substrate left, right, top, bottom & centre. The Dektak is a stylus profilometer that takes surface measurements using contact profilometry techniques.
- Atomic Force Microscope (AFM) measurements to confirm the depth, left, right, top, bottom, centre and selected grooves. The AFM can measure the depth of etch grooves that are less than 1nm deep.
Step 4
- Photolithography using Photo Lithographic Mask #3 and a negative resist.
- Ti and Al deposition, and lift off to form concentric circle and radial guidelines.
Step 5
- Photoresist deposition by positive resist. This layer is to protect the substrate from damage by quartz particles generated by diamond saw dicing.
Step 6
- Chip dicing, where a very fine diamond dicing saw cuts the substrate into individual chips, each 8 mm long by 4 mm wide.
Four chips with different etched groove depths are then assembled into the stainless steel slide, and covered with a coverslip using a UV activated adhesive. Current production has all Sets on the one larger chip.
Twelve major ‘proof of concept’ (POC) stages were necessary (POC1 to POC12), which included substantial development and testing, including distributing prototypes to the UK/HSE laboratories and to a number of NATA accredited laboratories.
B. Slide Assembly
Fitting tolerances are considerably less than 50µm, and assembly of such small components has to be conducted requiring considerable precision using a stereo-microscope. CAD designed jigs, fabricated on a 3D printer are used for most assembly stages to assist in maintaining this precision.
Further, all work conducted on the chips is undertaken in a custom designed laminar flow booth capable of ISO Class 3 (USA Class 1).
Modern UV curing acrylic adhesives are used, and cover slip placement is facilitated with a 3D printer converted to a ‘pick and place cover-slipping’ tool.
All stages of assembly are conducted in accordance with detailed protocols, coupled with extensive documentation to keep track of components and various measured performances during production.
Some Pickford Test Slides have some small but visible defects caused by the significant difficulty of using fused quartz in nanofabrication. Stringent quality control throughout manufacture and exacting certification protocols ensure they do not affect the function of the Slide in any way.
C. Slide Testing
During all stages of development of the Pickford Test Slide it has been critical to ensure that it was comparable to the HSE/NPL Mark II slide. To achieve this, several different approaches were investigated, as follows:
Subjective visibility is the gold standard to determine whether a Test Slide is acceptable or not, and is used on all assembled Pickford Test Slides to ensure that they are equivalent to the HSE/NPL Mark II Slides, and is also used by the UK HSE for Certification of all HSE/NPL Mark II Slides.
It became obvious however, that there was a need to develop a technique able to produce objective results, especially to be used during production.
Because the HSE/NPL Mark II slide design had been based on the phase shift of fine fibres, so was the initial design of the Pickford Test Slide.
However, the HSE/NPL Mark II phase shifts and visibilities do not line up with excellent work conducted by Rooker, and this approach was abandoned (Rooker, S. J., Vaughan, N. P., Le Guen, J. M. M., ‘On the visibility of fibers by phase contrast microscopy’, Am. Ind. Hyg. Ass. J. 43, 505-515, 1982).
A quantitative visibility measurement was developed using ‘grey levels’ (GLs), based on an 8-bit digital system (i.e. 256 individual levels), using a PCOM, a stable high quality microscope camera, and image processing software.
The image processing software chosen is Fiji, ‘Fiji is just ImageJ’, which is based on ImageJ, both open source.
A significant problem to overcome was the fact that the finer ridges and grooves moved out of perfect focus in terms of measurable grey levels when vertical stage travel was changed by 0.1µm – equivalent to approximately one tenth of a fine focus division – too fine for manual control. A solution was found by employing an external, mechanical focusing mechanism which increased the sensitivity by a factor of 11.
A specially developed Fiji macro is used to graph grey levels of every groove, in every Set of grooves. The macro then calculates a visibility factor based on the GL of the groove and the GL of the background. These factors are used in conjunction with subjective visibility tests as a matter of routine during the assembly of the Slides.
