Manufacturing Bits: Nov. 6

FISH metrology; locating atoms with AI; machine vision.


FISH metrology
The University of Illinois at Urbana-Champaign and the Mayo Clinic have developed a new molecular probe for use in imaging cells in living organisms.

The probe combines conventional fluorescence in situ hybridization (FISH) metrology techniques with compact quantum dots. This technology can measure and count ribonucleic acid (RNA) in cells and tissue without organic dyes.

RNA, which is a nucleic acid, is present in living cells. It carries instructions from DNA for controlling the synthesis of proteins. Some RNA also carry genetic information. Deoxyribonucleic acid (DNA) is a molecule that carries the genetic instructions in living things. A complete set of DNA is called the genome.

For years, meanwhile, the industry has used FISH metrology techniques. It images and counts DNA and RNA in single cells. The technology makes use of fluorescent dyes for imaging, but the dyes can deteriorate quickly. They also not good at imaging in three dimensions. Plus, it can only read out a couple of RNA or DNA sequences at a time, according to the University of Illinois at Urbana-Champaign and the Mayo Clinic.

In response, researchers from the two groups have enhanced FISH. They used quantum dots rather than fluorescent dyes. The quantum dots are made of a zinc, selenium, cadmium and mercury alloys. Quantum dots are produced using a chemical process. They are tunable and can convert short-wavelength light into colors spanning the visible spectrum.

Quantum dots illuminate the locations of individual mRNA as red dots in the cytoplasm of a single HeLa cell. The blue region is the nucleus. (Source: University of Illinois, Mayo Clinic)

The dots are 7nm in size and fit on a probe. “By replacing dyes with quantum dots, there are no stability issues whatsoever and we can count numerous RNAs with higher fidelity than before,” said Andrew Smith, an associate professor of bioengineering at the University of Illinois at Urbana-Champaign. “Moreover, we uncovered a fundamental limit to the size of a molecular label in cells, revealing new design rules for analysis in cells.”

Locating atoms with AI
The École polytechnique fédérale de Lausanne (EPFL) has combined machine learning and metrology to determine the location of atoms in powdered solids in record times.

The technology is ideal for the pharmaceutical industry. Many drugs are produced using powdered solids. Researchers would like to get a better understanding of their exact atomic-level structures.

In response, EPFL makes use of a technology called nuclear magnetic resonance (NMR) spectroscopy. It combines NMR with a machine-learning program. This can predict how atoms will respond to an applied magnetic field.

“Even for relatively simple molecules, this model is almost 10,000 times faster than existing methods, and the advantage grows tremendously when considering more complex compounds,” said Michele Ceriotti, head of the Laboratory of Computational Science and Modeling at EPFL’s School of Engineering. “To predict the NMR signature of a crystal with nearly 1,600 atoms, our technique – ShiftML – requires about six minutes; the same feat would have taken 16 years with conventional techniques.”

Machine vision
Imec has rolled out a new, high-resolution imager technology for use in industrial machine vision applications, such as photomask and wafer inspection systems.

The technology, called a time-delay-integration (TDI) imager, is based on a charge-coupled-device (CCD)-in-CMOS technology. The TDI sensors are manufactured on 200mm wafers by using a CCD process module inside a 130nm CMOS process flow.

The TDI imager has a quantum efficiency of more than 70% in the near-ultraviolet (UV) region, making it ideal for various machine vision systems. The new imager combines CCD TDI pixels and a CMOS readout technology in the same unit. This enables a low-noise, sensitive TDI performance with a fast and complex readout circuitry.

The UV sensitivity combines backside illumination technology with antireflection coating. With backside illumination, light enters the imager from the backside. This in turn boosts the efficiency of the image sensor.

Maarten Rosmeulen, program manager image sensors at Imec, said: “With this high-speed, low-power TDI imager, we are happy to expand our range of TDI CCD-in-CMOS image sensors with UV sensing capabilities. We can now offer TDI CCD-in-CMOS image sensors with high quantum efficiencies in both the UV and visible ranges, enabling a wealth of high-end applications, including remote sensing, medical imaging and industrial machine vision.”

Leave a Reply

(Note: This name will be displayed publicly)