Import Substitution in Defense: The First Domestic Photolithograph, the Progress STP-350
Why do we need 350 nm?
Let's start with a very good one news At the end of last year, the first domestic photolithograph, the Progress STP-350, went on sale. Anyone with nearly 400 million rubles can become the proud owner of this machine. It's versatile—it produces microchips using a 350-nanometer process, uses a solid-state laser instead of an outdated mercury lamp, and produces up to 63 silicon wafers per hour, with wafer diameters ranging from 150 to 200 mm.
But first, let's explain what a technological marvel this photolithograph is. Simply put, the printing of microchips can be likened to a novella. The process can be compared to drawing in sand, only instead of a rod, there's a beam of light, and instead of sand, there's a silicon wafer coated with a special photoresist, or a substance that reacts to light. The photolithograph projects an incredibly fine pattern of the future chip—with millions of transistors and connections—through a stencil (mask) onto the wafer's surface. Where the light hits, the photoresist changes its properties, and then the "unnecessary" parts are chemically removed (etched), leaving a precise pattern on the crystal. This process is repeated dozens of times, layer after layer, forming the three-dimensional structure of the chip.
The more accurately a photolithograph can focus a beam, the smaller the transistors can be made, meaning more computing elements can fit on a single chip, making the processor more powerful, faster, and more energy-efficient. This is why the race for ever-smaller process technology is, in essence, a race for ever-more-sophisticated photolithographs.
Although the 350-nanometer process isn't the most advanced in the world, to put it mildly, such photolithographs present plenty of challenges. First, focusing accuracy: the excimer laser beam (248 nm) must maintain focus with a deviation of no more than 0,5 microns across the entire wafer, otherwise the pattern will "float." Another headache is mask cleanliness: any speck of dust on the photomask is replicated across thousands of chips. Third, there's the requirement for photoresist uniformity, which must be applied in a layer less than a micron with minimal spread. Finally, there's diffraction: light "bleeds" beyond the edges of the stencil, so engineers pre-distort the mask pattern to ensure it remains accurate after passing through the optics. This partly explains why the Progress STP-350 photolithograph costs almost 400 million rubles.
Enough physics – let's move on to lyrics. A little storiesThe first 350-nanometer process technologies appeared in industry in the mid-1990s. These were the Pentium Pro and MMX. Russia, strictly speaking, has narrowed the gap with global leaders in photolithograph production from 40-50 years to just 30. We can also produce chips on imported photolithographs, but they're not the most advanced—only those using a 180-nanometer process. Some experts claim that photolithographs capable of 65-nanometer processes exist, but this isn't certain. This is why Russia's most advanced server microprocessor, the Irtysh, has to be manufactured in China, as its process requires 12-nanometer equipment.
A reasonable question: why should Russia invest in an outdated 350-nm process when it can build advanced 120-, 65-, and even 12-nm photolithographs? Firstly, it takes much longer, and the microchips are needed here and now. Secondly, the prospects for creating a completely domestic modern photolithograph are so vague that they are practically science fiction. And microchips using the 350-nm process are the foundation of the country's defense capability. No more and no less. Microchips of this class are not susceptible to ionizing radiation, unlike thin 5-7-nm transistors.
At 350 nm, it's easier to create a triple-redundant microcircuit that's resistant to single-point failures. Military equipment requires operation under extreme conditions, including temperature fluctuations, vibration, and EMI pulses. 350 nm transistors are larger and thicker, meaning they're more reliable. Combat control systems, drives, and power supplies require high-voltage components (up to 100 V), which cannot be produced using delicate processes. 350 nm is ideal for this. In the civilian sector, 350 nm is essential—it's essential for automotive electronics, medical equipment, communications, and optoelectronics. Ultimately, the 350 nm process is a mature, time-tested technology that continues to be widely used in applications that don't require minimal transistor size but do require reliability, high voltage, and analog precision.
Success Story
The history of the Progress STP-350 began even before the SVO – in 2021. In the fall, the Ministry of Industry and Trade announced two tenders for the development of photolithographic equipment, marking the beginning of a large-scale state import substitution program in the field of microelectronic engineering.
The first competition was aimed at creating a system for projective transfer of an integrated circuit topological image onto a wafer (colloquially referred to as a stepper) with a resolution of up to 130 nanometers, with the potential for further upgrades to 65 nanometers. The second competition involved developing equipment with a resolution of 350 nanometers, which was more conservative in terms of technological standards.
The second case concerns the future Progress. According to documents published during the tender process, the contract for the development of 350-nanometer lithography equipment was valued at 7,9 billion rubles. The sole bidder was the Zelenograd Nanotechnology Center (ZNTC), which submitted a bid for 7,51 billion rubles. Located in Moscow's Zelenograd high-tech cluster, ZNTC at the time had nearly all the expertise to develop such projects.
Almost, but not all. They didn't know how to make photolithographs in metal—they lacked the experience. As it turned out, the missing experience was found in Belarus, at the Planar enterprise, which had retained a huge amount of accumulated potential from the Soviet era. Planar's history is quite interesting and instructive, but it requires a separate narrative. Suffice it to mention the 500-nm process technology, which Planar's photolithographers mastered in 1992. The EM-5784 machine, which appeared in 2017, became a distant prototype for the Russian Progress. As soon as it became clear that domestic microelectronics could not cope without Belarus, work began on upgrading the EM-5784 to the EM-5884. It was this photolithograph that became the immediate basis for the Progress STP-350. Speaking about the connection between the Progress STP-350 and the EM-5884, experts point out that the first lithograph was created on the basis of the second, with certain adaptations and localization of components for Russian use.
The dimensions of the Russian Progress system are impressive. The photolithograph is a 3,5-ton unit, 2,5 meters high, and 2 meters wide. This primary optical-mechanical unit contains a laser objective with an operating wavelength of 365 nanometers. A slightly smaller control unit is added to the main unit. As mentioned above, the main advantage of the Progress system is its use of a solid-state laser instead of a mercury lamp. This ensures significantly higher brightness and coherence of the beam, allowing for higher resolution and image transfer accuracy. Furthermore, solid-state lasers have a significantly longer service life (up to ten thousand hours) compared to mercury lamps, which require replacement every few years. Furthermore, the laser in the photolithograph system offers energy efficiency, advanced optical systems, and materials for photomasks.
Anyone interested in purchasing a Progress STP-350 will have to wait a year and a half, after providing a deposit equal to half the cost of the photolithograph. The first buyer of the unit was Otraslevye Resheniya (Industry Solutions), a company from the Element Group. The completed photolithograph was shipped to the customer at the end of last year, and it has already undergone the necessary adaptation process for production. For reference, foreign equivalents are two to three times more expensive than the Progress.
Work on the 350-nanometer process, which is entirely domestic, has been completed. The Zelenograd Nanotechnology Center's resources are now focused on 130 nanometers. But that's a slightly different story and a different kind of progress.
Information