We recently spoke with Dr. Nicolaie “Mike” Moldovan, MEMS Lead Scientist at Advanced Diamond Technologies, about the company's microfabricated AFM diamond probes. These probes are based on discoveries at Argonne National Laboratory that resulted in a new form of diamond film called ultrananocrystalline diamond produced by chemical vapor deposition.
According to Dr. Moldovan, the company's process results in probes far superior in performance to those fabricated from silicon nitride, and at a price that is much more affordable than traditional diamond AFM probes with individually mounted and sharpened tips that can be priced as high as hundreds or thousands of dollars.
The annual market for AFM probes is estimated to be $50 million and the growth rate is healthy due to the proliferation of AFMs in biology, chemistry, mechanics and physics research labs as well as in industries such as semiconductor chip manufacturing.
MEMS Investor Journal: What is an AFM probe?
Dr. Moldovan: Atomic Force Microscopes (AFM) are instruments that probe a sample surface with an ultra-sharp tip (a few nanometers tip radius) reconstructing electronically an image with nanometer resolution allowing, with some enhancements, even atomic resolution. “Atomic force” means they are so sensitive that they actually can record forces as small as the interactions between individual atoms on the surface and at the tip, such as the Van der Waals force or ionic forces – the inter-molecular forces we hear about usually in chemistry – with nanoscale resolution. The tips used in these microscopes are mounted on microscopic cantilevers forming “the probes” - with different sizes, shapes, stiffnesses, resonant vibration frequencies, conductivities, optical, magnetic, or chemical properties, to enable the many operating modes already developed for such microscopes. They resemble the diamond styli used with vinyl records although AFM probes are much smaller and extremely sharp. In addition to imaging applications, the probes are used for manipulation of materials at the atomic/molecular level and also in nanomanufacturing.
Atomic force microscope basic principles. Image courtesy of AFM manufacturer Asylum Research, Santa Barbara, CA.
MEMS Investor Journal: Where are AFM probes used? What kinds of companies are buyers and users?
Dr. Moldovan: AFMs are used everywhere one needs to see details on the nanoscale. Examples include most biology, chemistry, mechanics and physics research labs as well as industries such as semiconductor chip manufacturing. If optical microscopes are already components of high school labs, with the development of nanotechnology AFMs will eventually become ubiquitous devices in that environment too.
Recently, companies such as IBM and Nanochip began to use AFM-type probes to record data in polymers. This innovation will result in high density non-volatile memory devices that will cause the use of probes to increase by many orders of magnitude. NanoInk uses AFM probes to “write” molecules using organic inks. Applications such as those being developed by IBM and Nanochip will put new demands on such probes – for example, the probes will need to last much longer.
Image of an Atomic Force Microscope (left) and images of Advanced Diamond's diamond AFM probes (middle, right).
MEMS Investor Journal: What is the overall size of these markets in terms of dollars? Upon what assumptions are these estimates based?
Dr. Moldovan: AFM probes are consumables. We’ve looked high and low for reliable market numbers and the estimates are wildly divergent, ranging from $25 million to about $100 million per year. We believe that $50 million is about the median based on the polling we’ve done although everyone agrees that the growth rates are pretty healthy.
The market is growing due to the proliferation of atomic force microscopes and new applications like the ones described above. In addition, we expect the market to expand considerably due to the availability of our diamond probes since one of the main impediments to wider adoption of AFM, especially for nanomanufacturing, has been their poor durability. The need to replenish probes often creates much inefficiency and adds considerably to the cost of using AFM.
MEMS Investor Journal: Typically, how long do probes last compared to what can be expected for your diamond probes?
Dr. Moldovan: The performance is very dependent on the application. For instance, we have performed laborious tests to compare the imaging stability of diamond probes and standard SiN probes during contact mode scanning of a diamond substrate in normal lab air during which no load is applied to the tip.
It is important to note that zero net load on an AFM probe actually means some of the tip atoms, closer to the sample, are strongly repelled, while others, farther away, are attracted to the sample. The combination leads to a particular distribution of stresses in the tip at the nano-scale, which is responsible for wear even at zero net force. This is the normal mode used when imaging a surface in contact mode. In these conditions the relative performance is dramatically different, with the SiN probe degrading almost immediately whereas our diamond probes show no change even after 100 scans.
The tip durability graph below shows the evolution of the tip radius for our probes and a SiN probe with the number of scans. One scan corresponds to 512 mm of scanning distance, the length necessary to cover an area of 1 mm times 1 mm of imaged field. While the SiN tip degrades steadily and significantly from about 10 nm radius to about 80 nm radius in 60 scans, the diamond tip shows a stable radius over the 100 scans performed. Note also the large fluctuations in the tip radius for the SiN probe, showing that pieces of material are randomly removed or attached to the tip, rendering its shape unstable and unreliable. Under these zero load imaging conditions, the wear of our probe is insignificant and cannot even be numerically compared with SiN probes.
Changing the conditions of the test by either applying a higher load during scanning or changing the relative humidity from 50% to 10% can have a big impact. We went to higher applied loads and scanning speeds as an attempt to observe some measureable wear on our probes, but even with as much as 280 nN of force applied over a total scan distance of 220 mm there was still no wear that was measurable within the error of the measurements.
Ultrananocrystalline diamond probes vs. SiN probes. Source: R. W. Carpick, University of Pennsylvania.
MEMS Investor Journal: Who are the main players in the AFM market and what is their market share?
Dr. Moldovan: The main producer of AFM probes is Nanoworld and, according to one market report, it has more than 45% of the market. Veeco Instruments, based in Plainview, NY, is the world’s largest manufacturer of atomic force microscopes and has a large probe business.
MEMS Investor Journal: What are ADT’s diamond probes and how do they differ from other types being used?
Dr. Moldovan: Most AFM probes today are made out of silicon (Si). Where extra durability is required for contact mode imaging, people often choose silicon nitride (SiN) probes. The lifetime of AFM probes is limited by the fact that such probes wear during scanning the sample surface. This is analogous to a pencil point getting blunter as it is used.
Our AFM probes are made from a form of nanocrystalline diamond we call UNCD. Diamond, the hardest material available, assures the lowest wear rate as shown in the tip durability graph above.
The use of diamond for AFM probes itself is not new. Traditionally, diamond probes were produced as individually-sharpened diamond pieces mounted on cantilevers with glue, or as batch-processed diamond-coated silicon probes. The former are very expensive and bulky, while the later are not sharp enough – their tip radius is equal to the thickness of the diamond coating which is about 100 nm. Moreover, the consistency of diamond-coated probes is notoriously poor.
Our diamond probes are entirely batch processed and not limited in terms of sharpness. We have manufactured probes as sharp as 4 nm in tip radius.
MEMS Investor Journal: What would lead to the selection of your diamond probes versus conventional probes?
Dr. Moldovan: The changeover of a probe is a time consuming and disruptive process. Tip wear shows up as a loss of resolution and gradual worsening of the imaging performance. For silicon probes the chemistry of the probe tip is also not stable, especially when imaging in normal lab air. If the AFM is used as a metrology tool (such as for characterizing the roughness or nanostructures), the degradation of the tip leads to inaccurate measurements. If the same imaging performance can be achieved with diamond probes that have better price performance because they last so much longer, most AFM users would be happy to switch. For the same tip radius, a diamond probe deforms less at the contact point then other probes, thus allowing more accurate imaging. Diamond probes maintain their overall dimensional stability, which means they can be used for a larger number of measurements before changeover. More durable probes also mean that larger areas can be imaged.
Diamond can also be made highly electrically conductive, meaning that probes can be made that exhibit very low contact resistances on conducting surfaces. There are a number of specialty applications and imaging techniques that leverage the electrical conductivity of the probe such as scanning spreading resistance microscopy (SSRM) and electrochemical imaging (ECAFM).
Aside from unsurpassed hardness, diamond is also a chemically inert material. A big problem with Si probes is that Si is a reactive material and biological fluids stick to it. We believe that the inertness of diamond is an attribute that is of great value to many users - especially those imaging in aqueous media. Additionally, diamond probes extend the application range of AFM probes. This supports tip-based nanoscale mechanical measurements, nanomanufacturing, and lithography.
Another figure of merit for AFM probes is the frequency response of the cantilever. Because diamond is the world’s stiffest material (high modulus) it vibrates at a much higher resonant frequency than other materials for a given probe geometry and also with a higher quality factor in air. This enables dynamic mode imaging at standard high frequencies today (300-500 kHz) and will enable the development of rapid imaging techniques based on probes that operate at frequencies greater than 1-2 MHz.
MEMS Investor Journal: Why is the higher frequency significant?
Dr. Moldovan: An increased frequency of probing the surface results in the possibility to scan faster or image larger areas in shorter times with the same spatial resolution. Certain dynamic, non-contact mode operation methods of AFMs make them sense the atomic forces by a change in vibration frequency of the cantilevers while oscillating above the sample; the force-sensitivity in this case increases with the vibration frequency and the quality factor, leading to better imaging capabilities.
MEMS Investor Journal: Is there anything holding back the broad adoption of these probes given their superior price/performance advantages over conventional probes?
Dr. Moldovan: Nothing is holding us back. We introduced contact mode AFM probes last September and have just now started to market them aggressively. We have signed agreements with major distributors, and we expect to announce more with several market leaders shortly. In addition we will bring to market several new probes for dynamic mode imaging and ones based on electrically conducting diamond in the next few months. The only major holdup is that AFM users are used to probes that either fail quickly or don’t work at all—or have tried diamond-coated probes and been very disappointed with their performance—that they stick to the “status quo”. It’s pretty clear in our interactions that, if at universities the buying decisions were up to the graduate students using the AFM that we would be deluged with orders.
MEMS Investor Journal: What applications are not suitable for your diamond probes?
Dr. Moldovan: There are applications and usage environments in which frequent changeover of probes is required for reasons other than wear. Probing highly contaminated samples may be an example. Imaging of high-aspect ratio structures such as step profiles may not be suitable for the current generation of our probes, but plans exist to bring HARD (high aspect ratio diamond) probes to market within the next year.
MEMS Investor Journal: What led ADT to apply MEMS technology to these probes? What was “wrong” or “unsatisfactory” or “could be done better” vs. conventional AFM probes?
Dr. Moldovan: Si and SiN AFM probes use MEMS technology in their fabrication. This allows manufacturers to take advantage of wafer-scale manufacturing to make affordable probes. Traditional diamond AFM probes with individually mounted and sharpened tips are priced in the hundreds or thousands of dollars range because they are made one at a time. Using traditional MEMS technology to make diamond probes was the only option to make them affordable. This became possible with our proprietary diamond deposition process and material which captures the unique attributes of nanocrystalline diamond including smoothness and low film stress.
We should also mention that traditional diamond films, such as those used to coat cutting tools like drill bits, are too rough and have much too high film stress to be used for MEMS applications. Nanocrystalline diamond uniquely enables this capability, and AFM probes are but the first tangible application of diamond MEMS.
MEMS Investor Journal: What is unique about your solution?
Dr. Moldovan: Our diamond AFM probes are made essentially the same way as SiN probes are made except that instead of SiN we are using nanocrystalline diamond on which we have several patents. As has been mentioned, our nanocrystalline diamond composition has advantages over other types of diamond film, including the fact that it is very smooth. The key relies in the deposition process, leading to extremely small grain sizes – 2-5 nanometers, with no non-diamond phases at the grain boundaries (see TEM image below). The composition also allows the tuning of properties, including electrical and thermal conductivity, which is beneficial to a large variety of diamond probes. Today we have just a few varieties of probes in the market, but we will be offering more varieties soon.
Transmission Electron Microscope (TEM) image showing the grain sizes in ultrananocrystalline diamond, which is the key to the smooth yet hard diamond technology.
MEMS Investor Journal: What design challenges did you have to overcome in bringing your solution to market and how did you overcome them?
Dr. Moldovan: Developing probes that have tip radii less than 20 nm required a careful optimization of the deposition process of ultrananocrystalline diamond to ensure the complete filling of the tip apex. This required difficult and costly high-resolution scanning and transmission electron microscopy to evaluate the shapes and structures of the tips during many deposition trials. Since in our process the tips are fabricated on a silicon wafer with the tip facing towards the silicon, a second challenge was to bond the diamond layer onto a glass substrate and then remove the silicon. A good solution for this bonding had to be found and optimized, since nanocrystalline diamond does not bond directly to glass with the technology used for the fabrication of silicon nitride probes. We found an interlayer with good adhesion to both components that allowed this bonding.
MEMS Investor Journal: How are your products being marketed?
Dr. Moldovan: Our ultrananocrystalline diamond probes are currently available directly through ADT via our website www.thindiamond.com and through Nanoscience Instruments globally and ITB Corporation in South Korea.
Our strategy is to come out with more varieties of diamond probes including those suitable for dynamic mode imaging and those made of conducting diamond. We will add new distributors to provide more availability, and will be announcing some strategic alliances in the coming months.
We are very receptive to hearing from power AFM users who may have new ideas for how our diamond probes can be used; or from those users who are willing to publicly share their stories of their successes.
MEMS Investor Journal: What are your price points and how do they compare with conventional probes?
Dr. Moldovan: The prices for our diamond probes can be easily found on our website or through our distributors. We have purposely priced them to be much more economical than very expensive diamond-tipped probes (upwards of $1,000 a probe) and comparable to diamond-coated Si probes ($60-$120 each), which, while relatively inexpensive, are notoriously inconsistent in performance. In our own tests our diamond probes showed a performance/price improvement over SiN probes of more than 30 times while being priced at approximately $125 each.
MEMS Investor Journal: How are your diamond probes being received in the market?
Dr. Moldovan: Our probes have only been on the market for a few months and initial reception has been very encouraging. Moreover, customers are looking for additional varieties such as dynamic mode and conductive diamond. We are working to expand these products into our distribution channels.
MEMS Investor Journal: What do you see as future applications? Where is your probe technology going?
Dr. Moldovan: We are quickly developing new varieties of diamond probes prioritized around those applications where diamond makes a compelling difference. Such fields include metrology probes and conductive probes for electrical and electrochemical scanning modes, and high-aspect ratio probes. In the long run, we envision more complex probes, with integrated functionalities such as actuation and sensing, and probe microsystems where the wear resistance and superior life time of diamond probes adds to their performance in a manner that is meaningful to customers.
Our manufacturing infrastructure is in place and we have overcome the technical challenges to making diamond probes. We fully expect to be able to come out with new varieties at a quick pace.
******************************************************
******************************************************
Nicolaie ("Mike") Moldovan is a top micro-nanofabrication expert with 20+ years of experience in materials, process, and device integration. He joined ADT after being a research professor with Northwestern University where he developed diamond AFM probes, nanofountain probes for AFM nanolithography, and carbon nanotube devices. During 1998-2002, with Argonne National Laboratory, he conducted research in ultra-deep x-ray lithography, x-ray micro-optics, field emitters, and ultrananocrystalline diamond microfabrication technologies. From 1983 to 1998 he developed electronic devices, micro-optics, and microfabrication technologies at the Institute for Electronic Components and the Institute of Microtechnology in Bucharest, Romania. As head of the Laboratory for Unconventional Microfabrication Technologies, he ran international projects in LIGA technology, silicon, germanium, metal, and glass micromachining for micromechanics and micro-optics.
Dr. Moldovan received his M.Sc. degree (1982) and his Ph.D. (1997) in Physics from the University of Bucharest, Romania. He completed two DAAD research stages in the Fraunhofer Institute for Solid State Technology in Germany, where he specialized in anisotropic etching of silicon and germanium, and stress engineering in thin films and membranes. Among his achievements is listed a master equation-based atomic scale model able to reproduce the Si-KOH etching anisotropy diagram starting from atomic and lattice parameters. He has authored more than 100 scientific papers, five patents, and received an R&D 100 award.
Copyright 2009 MEMS Investor Journal
Comments