Stable integrated ultra-pure gas supply system for semiconductor manufacturing

2.1 Innovation

Diagram 1 shows the stable integrated ultra-pure gas supply system for semiconductor manufacturing

Diagram 1. FCS_® equipped Integrated Gas System (IGS_®)

Development started with the development of the UPG_® fittings, and then the W-Seal was adopted for the construction of the IGS_®. Concurrently, the FCS_® and the ECV_® (not pictured) with a fast valve open/close response speed of a few milliseconds were developed.

Diagram 2. Conventional gas system equipped with MFC

Conventional gas systems are constructed by connecting single devices such as pressure regulators, MFC and valves together with pipes and fittings. Due to this, it is necessary to loosen the support of the surrounding devices when replacing a single device during maintenance, which takes time. As a result of integration, maintenance can be done regardless of other devices by just loosening the W-Seal bolt of a single device, which easily shortens the required time. Also, since the integration does not require piping or fittings, it is a third smaller when compared to a similar system (Diagram 3).

Diagram 3. The IGS_® is a third of the size when compared to a conventional system with the same flow

We performed simultaneous experiments where we read the output signals and actual flow rate signals when pressure flucuations were applied to the upstream of the FCS_® and MFC flow rate controllers. The results were that the FCS_® was not affected at all by the upstream pressure fluctuations (Diagram 4). On the other hand, due to the upstream pressure fluctuations having an effect on the MFC, installation of pressure regulators and pressure guages on the upstream are essential.

Diagram 4. Difference in upstream pressure for flow rate controllers

Accordingly, a FCS_® equipped gas system does not require pressure regulators and pressure guage monitors, which further reduces costs and size.

The ECV_® valve has a FCS_® orifice in the seat section, which allows the unification of both, and creates a control valve to counter transient response (Diagram 5).

With the FCS_®, the pressure of the upstream orifice is constantly maintained and controlled, and after inputting the flow rate control signal into the FCS_®, the gas can be controlled by just opening and closing the ECV_® (Diagram 6).

Diagram 5. Flow rate control valve with transient response countermeasures (left) / Diagram 6. ECV_® flow rate characteristics when opened (right)

In order to confirm the gas supply with a water supply system that uses a FCS_® equipped IGS_® that takes into account transient response characteristics, diagram 7 shows the confirmed results and characteristics taken from semiconductor manufacturing equipment at Tohoku University.

According to Professor Ohmi, "The long dream of an ideal gas supply system that supplies gas with a water supply system has finally been achieved."

Diagram 7. Process gas concentration and variations over time

Details for each technology are shown below.

2.1.1 IGS_® innovation

• ① Inheriting the concept of UPG_® fittings, it uses the most reliable and quality W-Seal fitting.
• ② To stabilize the sealing between blocks, the fixing method for the base block has a leveling system.
• ③ The interface uses two bolts for one seal point, which stabilizes the sealing and makes flow design easy.
• ④ The diaphragm valve used in the IGS_® has a special shape for the flow channel, which contributes to standardization and makes it possible to design for flows required by several types of valves. Also, in the semiconductor manufacturing process there are concerns with the purity of the gas due to the accumulation of gas in the gas switch location (dead space), but this valve has a unique shape which has no dead space at all.

With the above innovations, many equipment and semiconductor manufacturers have adopted the IGS_®.

2.1.2 What are UPG_® fittings?

The UPG_® (Ultra compact metal gasket fittings) has been devised to dramatically improve performance with a newly designed seal construction for metal gasket fittings, which have been widely used as fittings for semiconductors. With this compact feature, there is no internal dead space (accumulation of fluids), and the reliability of the seal is improved, and the reliability of the pipe construction (Diagrams 8, 9, 10).

Diagram 8. Seal for UPG_® and conventional fitting

Diagram 9. UPG_® and conventional fitting

Diagram 10. UPG_®

• 1. 70% the size of the conventional fitting
• 2. No dead space (fluid accumulation)
• 3. Gasket material is SUS316L double melt (double layer melted material and clean stainless steel) and has high seal performance.
• 4. Can be constructed with low torque
• 5. After tightening, the gasket comes into contact with the whole surface and handles the external pressure, with a highly reliable seal.

Conventional fittings (metal gasket fittings) have a seal construction that is sealed in one location, but the UPG_® (compact metal gasket fitting) has a construction that is sealed with a 3 stage process.

With this 3 stage seal constrcution the UPG_® fitting has:

• 1. Improved sealing
• 2. Improved strength against external force
• 3. As the contact surface changes during construction, you can feel the tightening as you tighten it, which makes it possible to safely construct it.

Diagram 11 shows data for the seal performance of the UPG_®. The horizontal axis shows the tightening torque, and the vertical axis shows the leak amount. The specified torque (torque for tightening) is 10.8 N⋅m (110 kgf⋅cm), but the leaks stop at 3.9 N⋅m (40 kgf⋅cm). The sealability is good and there is margin in the seal, which makes it highly reliable against leaks.

Diagram 11. Relative torque and leak amount

Diagram 12 shows the change in tightening torque during construction of the UPG_®. As you increase the tightening torque for conventional fittings, the tightening angle changes linearly with it. With this, it is difficult to judge how much it should be tightened. With the UPG_®, the contact area of the seal surface changes, which means the tightening angle does not change linearly as the tightening torque increases, and the angle changes as you tighten it. As a result, the constructor will get a heavy feeling as they tighten and they can accurately judge the tightening, which allows safe construction.

Diagram 12. Relative torque and leak amount

In order to position the piping after constructing the fitting, the pipe is twisted which causes leaks. Diagram 13 shows the assessment of the torsion after pipe construction for the UPG_® and conventional fitting. With the conventional fitting, as you twist, leaks occur at the same time, but with the UPG_®, leaks do not occur unless it is twisted more than 70゜. This also indicates that the margin for the seal is high (Diagram 13).

Diagram 13. Torsion and leak test

Inheriting these characteristics of the UPG_®, the W-Seal uses a bolt connection instead of a union nut (Diagram 14).

Diagram 14. W-Seal

2.1.3 W-Seal characteristics

• ・By adopting the same seal method as the UPG_®, the ultimate leak amount can be achieved. This is less than one quadrillionth of a cubic centimetre (It would take more than 30 million years to leak 1 cc.).
• ・Dead space free and particle free.
• ・A leak check hole is provided to easily inspect for leaks in the seal.
• ・Can be attached/detached more than 100 times.
• ・Seal construction is done by just tightening two bolts.
• ・Can be sealed with low torque (4.9 N·m (50 kgf·cm)(φ6.35 size)), improving workability.

Diagram 15 shows data for the W-Seal seal performance. The horizontal axis shows the tightening torque and the vertical axis shows the leak amount.

The specified torque is 4.9 N·m (50 kgf·cm), but the leaks stop at 0.8 N·m (9 kgf·cm). The sealability is good and there is margin in the seal, which makes it highly reliable against leaks.

Diagram 15. W-Seal leak test

Also, after constructing the W-Seal, we checked the leak performance when external force is applied, and the results are shown in diagram 16. Even when external force is applied, there are no leaks, which is very safe.

Diagram 16. W-Seal external leak test

2.1.4 Leveling system

To construct the IGS_® components, place the lower block fittings on top of the metal plate, and then place the other devices such as the valves and MFC and top. The flow channel for the gas flow alternates between the lower block fittings and the upper devices, but due to fluctuations in manufacturing the metal plate (dimensions, undulations, etc.) and the lower block (height dimensions, etc.), the upper device joint (W-Seal) is affected. The leveling system is a system that absorbs the fluctuations in the metal plate and lower base block. The leveling system is a system where a rubber washer is installed inside the lower block, which moves the lower base block up and down with the rubber's elasticity. When connecting the upper devices, the lower block naturally follows the seal surface of the upper device, which doesn't hinder the W-Seal seal and stable seal performance is achieved.

2.1.5 W-Seal IGS_{® has} interface design (Diagram 17)

Diagram 17. IGS_® design

The W-Seal IGS_® interface design uses two bolts for one seal (valve inlet and outlet have separate seals), but interfaces from other companies use four bolts for two concurrent seals (valve inlet and outlet are sealed at the same time).

The W-Seal IGS_® design makes it possible to reduce the number of connection points and have a flexible flow design. The diagram shows that the W-Seal design seals require only eleven points, while the other designs require sizteen points for the same flow.

2.1.6 ECV_® innovation

The ECV_® opens and closes the gap between the valve seat and guard to increase pressure by using the power of the coil and the vertical motion of a cylindrical shaft unified with a magnetic component which is driven by electromagnetic induction in the coil. Compared to a general pneumatic valve which has to replenish and release gas to operate, the time required to open and close the valve is around 1/20 of milliseconds. Also, by developing the magnetic component to use iron and cobalt or iron and nickel alloys with a saturation magnetic flux density above 2 T (Tesla), the magnetic component is much smaller and the flow control valve is also made more compact. An initial large electric current is used to supply the excitation current to the coil and open the valve, but to maintain the open state, only a small electric current is used. By separating these two supplied electric currents, the energy consumption of the coil is suppressed to a small value.

2.1.7 FCS_® innovation

The FCS_® (Diagram 18) is the semiconductor industry's first pressure controlled flow controller that was developed by Fujikin. Flow rate control of gas in semiconductor manufacturing has been done by mass flow controllers (MFC) developed in America for a long time. The reasons why FCS_® is being adopted in place of MFC are:

• ① There is no flow rate fluctuation due to the supply pressure fluctuation.
• ② It is very precise.
• ③ Differences in flow rate are small.
• ④ It has a self-diagnosis function.
• ⑤ Does not depend on gas type.
• ⑥ No pressure regulators are required, which reduces costs.
• ⑦ IGS_® can be made ore compact.

Diagram 18. FCS_®

2.1.8 Construction and principles of the FCS_® (Diagram 19)

Diagram 19. Comparison of FCS_® and MFC control systems

We will first explain the MFC. The thermal sensor in the MFC has a bridge circuit constructed with two coils wound around a small tube. As the coil temperature increases, gas is flowed into the small tube and the temperature of the coil changes with heat transfer. The bridge circuit detects the temperature change as a change in resistance and controls the flow rate. Since the thermal sensor is a small tube, a large gas flow cannot pass through it, so a small amount of gas from the laminar flow element is bypassed and passed through the small tube. The main components are the thermal sensor, laminar flow element, flow control valve installed on the downstream, and an electric control circuit. In order to have a fixed mass flow rate, the control valve's lift is controlled by the electric circuit. However, the repsonse speed is slow due to the thermal sensor, and it is difficult to supress slow rate fluctuations due to pressure fluctuations.

The FCS_® is based on critical expansion conditions. The critical expansion condition is when the orifice upstream's absolute pressure is around two times higher than the downstream's absolute pressure, and the gas is passed though at sonic speed. The flow rate is dependant on only the upstream pressure, and the flow rate is relative to the pressure. For example, if the downstream pressure is 1 kPaabs., and the upstream pressure is above 2 kPa abs., the critical expansion condition is created. If you use an orifice that has a flow rate of 10 cc when the upstream pressure is 3 kPaabs., then at 30 kPaabs. the flow rate will be 100 cc, and at 300 kPaabs., the flow rate will be 1000 cc. The FCS_® is a single device that integrates this principle. Since semiconductor processes have a decompression process using a vacuum, it is possible to the critical expansion condition.

The main components for the FCS_® are a piezo control valve on the upstream, pressure sensor, a digital control circuit alongside the orifice, and a temperature sensor installed on the main body. With the MFC, the control valve is positioned on the downstream side of the thermal sensor, but the FCS_® has the control valve positioned on the upstream side of the pressure sensor.

When the gas supply pressure is at 250 kPaG (351.3 kPaabs.), the pressure is maintained at 300 kPaabs. and the flow rate is adjusted to be 100%. Therefore, a flow rate of 50% is achieved with a control pressure of 150 kPaabs., and 10% is achieved with 30 kPaabs. Depending on the flow rate range you want to control, choose an with a hole diameter of around 10 μm to 500 μm.

2.1.9 FCS_® characteristics

① Flow rate does not fluctuate due to supply pressure fluctuations.

The flow rate for the FCS_® is controlled by the pressure on the upstream side of the orifice. Even if the supply pressure fluctuates, the pressure sensor and the piezo control valve have a response time of less than 1 ms, and with quick feedback control, a fixed pressure can be maintained. It has been confirmed that supply pressure fluctuations have no affect at all.

② It is very precise. ③ Individual flow rate differences are small.

In the semiconductor industry, the gas flow rate is indicated by the volumetric flow rate when it has been converted to standard conditions (0℃, 1 atm).

The guaranteed precision for the FCS_® flow rate is within ±1% set point. In order to achieve this, the zero point temeprature drift for the diffusion semiconductor pressure sensor is corrected by a digital calculation.

The reason why this digital pressure guage is more precise than any other pressure sensor sold as a separate unit is because of this zero point drift correction. In addition, the flow rate adjustment uses the traceable flow rate reference device at the National Metrology Institute of Japan, and with digital correction of the calcualted pressure, ±1% set point is guaranteed.

④ It has a self-diagnosis function.

When the orifice diameter is changed by corrosion or clogged up with foreign substances, the FCS_® cannot control the flow rate correctly. For this reason, we developed a function that can self-diagnose abnormalities in the orifice. When the control valve is closed from a 100% control state, the pressure drop characteristics during the escape of residual gas in the space between the orifice and the control valve is forwarded to memory, where comparisons can be made to detect any abnormalities in the orifice. If the maximum flow rate deviates by more than 3%, the system sends an abnormality signal.

Recently, 300 mm wafers are used for semiconductor manufacturing, and it is commonplace to process each wafer one at a time. The semiconductor product value that can be taken from one wafer is said to be from several millions to tens of millions of yen. Therefore, in order to check if the process is correct each time one wafer is processed, a flow rate self-diagnosis function is used in the inspection system.

⑤ Not dependant on the gas type.

The theoretical flow rate formula (Diagram 20) for the gas passing through the orifice with the pressure conditions created by the critical expansion condition, is the product of the orifice pressure multiplied by the orifice cross section area, divided by the square root of the orifice gas temperature, which is then multiplied by the product of the gas properties such as the heat capacity ratio, gas constant, molecular mass, and gas density, etc. Accordingly, after calculating the gas properties with the theoretical values of each gas, we get the flow factor of the gas which is proportional to the nitrogen value of 1. This means that with the same pressure, orifice and temperature, except for nitrogen, the gas flow rate is gained by multiplying the flow factor of the measured flow rate of nitrogen gas.

Diagram 20. Theoretical flow rate formula for gas passing through the orifice under critical expansion conditions

When we took actual measurements of more than 40 types of special gases, most gases were within ±1% of the calculated flow rate multiplied by the flow factor. For example, HCI (hydrogen chloride gas) has a flow factor of 0.870299, and He (helium gas) has a flow factor of 2.808319. With the FCS_® adjusted to 1000 sccm for nitrogen gas, HCL is 870 sccm, and He is 2808 sccm. The flow rate adjustment for the FCS_® is all done by nitrogen gas, and when the customer uses it, it is multiplied by the flow factor and converted to the actual gas flow rate. The conventional MFC was individually adjusted by each gas type and flow rate. Therefore, in order to prepare for unexpected circumstances, spare parts for every gas type and flow rate for the MFC being used were stored in inventory at the semiconductor manufacturing plant. At the least this exceeded a few hundred parts. On the other hand, all FCS_® systems are calibrated by nitrogen, so if you apply the flow factor to any gas, the range of gases used is decided, and if you store ten devices for different flow rates in inventory, you can support any gas and any flow rate, which completely changed the idea of spare parts in inventory.

Diagram 21. List of flow factors (F.F.) for each gas

⑥ Pressure regulator is not required so it can be more compact and reduce costs

For semiconductor manufacturers, productivity and large cost reductions have been made by refining and increasing the size of the wafers, and the degree of integration is improved. However, they also simultaneously demand cost reductions for semiconductor equipment and components. By adopting the FCS_®, pressure regulators and pressure sensors for monitoring are not required since supply pressure fluctuations do not affect it, which contributes to cost reductions. In addition, it is more compact, which has reduced the area required in clean rooms which have high maintenance fees.

Diagram 22. IGS_® using a FCS_®

2.2.1 IGS_® technology level

The lower base block and the upper devices are joined together by bolts, but the clamping force is decided by the thrust of the bolt. Since semiconductor equipment needs to be completely oil free, the blocks and the bolts also need to be completely oil free. Since galling (metal adhesion) occurs when oil free metal is used for joining, the bolt used in the IGS_® (Diagram 23) has a special silver plating treatment applied to it. Although silver plating is mainly used for decoration, the silver plating used on this bolt has a special quality that allows it to have a stable coefficient of friction.

Diagram 23. IGS_® silver plated bolt

The body side of the W-Seal seal part (base, device: Diagram 24) and the inserted gasket use the same SUS316L material. The fitting must be durable against attachment/detachment, so the W-Seal seal front has been specially hardened (Since corrosion resistance is reduced with hardening treatments such has heat treatment, etc., a mechanical hardening treatment is used.), and it does not deform at all when attaching and detaching. Normal stainless steel has a hardness of 200 HV, but the W-Seal seal front has 380 HV. On the other hand, the gasket which deforms for the seal is 100 HV with annealing. With normal stainless steel, it only lowers to 150 HV with annealing, but the gasket in the W-Seal uses a double melt material and heat treatment conditions that accomplish 100 HV. With a hardness difference of 280 HV between the body seal front and the gasket, attachment and detachment of more than 100 times is achieved.

Diagram 24. Device and base block for the IGS_®

Furthermore, in order to position the device and protect the seal surface, the gasket is combine with a guide ring (Diagram 25). When the device is installed into the base block, the gasket is installed into the device and the outer side of the guide ring makes it possible to smoothly install the device into the correct position. Also, since the gasket's seal surface is on the inside of the guide ring, the seal surface is difficult to damage.

Diagram 25. Gasket with guide ring for the IGS_®

2.2.2 ECV_® technology level

Actuators for clean specification valves use pneumatic valves due to the superior thrust they generate, and compact clean specification electric valves have not really made an appearance on the market yet. However, by developing a magnetic body with iron and cobalt, or iron and nickel alloys with a saturation magnetic flux density of more than 2T (Tesla), the magnetic body can be made much smaller and it is possible to develop a compact clean specification electric valve.

A large initial electric current is used to open the valve, and a smaller electric current is used to maintain the open state. With this separation, it is possible to prevent temperature increase in the coil, and lower energy consumption when the open state is being maintained.

The valve response time became faster, and since the number of times the valve opens and closes is increased, durability for the open/close lifespan needs to be higher than pneumatic valves. The ECV_® has achieved an open/close lifespan of more than ten million times.

2.2.3 FCS_® technology level

The orifice (Diagram 26) uses the triple melt material SUS316L, which reduces impurities such as sulfur and non-metallic inclusions in aluminum as much as possible. A 10μm to 500μm hole is drilled into the 50μm thin plate. Finally, electropolishing is performed on the interior of the hole to make it ultra clean. Since it is very difficult to handle a 50μm orifice metal plate as it is, in order to fix it to a part created from a round bar, we initially used laser beam welding technology. To meet the demands for weldless construction, we also used mechanical fasteners. In order to inspect if the optimal hole to guarantee the specified flow rate of the orifice has been completed, we initially used a scanning electron microscope​, but it was impossible to check with a tolerance of ±1μm. For this reason, the pressure for the upstream of the orifice is fixed, and we developed a dedicated device for measuring the flow rate. We used this flow rate inspection as a way to inspect the hole diameter of the orifice.

Diagram 26. SEM photograph of the orifice shape (1000x magnification)

The gas connection part for the pressure sensor uses FS9, a super ferrite material. After machining the diaphragm, it is thinned down to 60μm with a lapping process, and then subjected to Cr203 passivation treatment developed by Professor Ohmi at Tohoku University, while taking in account corrosion resistance, fluid loss and noncatalytic characteristics. Since variations in the pressure sensor's diaphragm greatly affects the sensor's characteristics, it is not easy to give it a mirror finish while controlling the variations in thickness. This diaphragm is supplied to pressure sensor manufacturers and a dedicated pressure sensor to the FCS_® is made. In addition, the pressure sensor is sealed against external leaks with a metal gasket.

It is unheard of that a pressure sensor which is extremely sensitive to deformations, is sealed with a metal gasket.

Since we could only use a metal gasket to guarantee reliability against external leaks, with the knowledge we have accumulated, we planned everything for the fastening, metal gasket design, manufacturing, aging/correction of the fastening, and accomplished our goal.

Flow rate reference devices are not sold. Everyone knows that gas volume changes according to the temperature and pressure. For this reason, standard conditions (0℃, 1 atm) are used for the volumetric flow rate of the gas in the semiconductor industry. At the National Metrology Institute of Japan, the gas flow rate is calculated by measuring the weight of the gas every unit of time and then converted to the volumetric flow rate.

The Chemicals Evaluation and Research Institute, Japan received the technology from the National Metrology Institute of Japan, and carried out flow rate correction research on the MFC. Therefore, it is inevitable that the MFC is used as a reference device. Using the calibrated MFC, the mass flow meter (MFM) with no flow rate adjustment function was calibrated with the same theory as the MFC, and used in the company as a flow rate reference device. In order to match the flow rate to the flow rate reference device, a dedicated automatic flow rate adjustment device was manufactured. Although flow rate calibration was performed with the FCS_® installed on the downstream of the MFM, it was extremely difficult due to the output of the MFM fluctuating with a ±1℃ temperature difference in the clean room. We discovered that touching the pipes with a hand disrupted the output.

In particular, it was impossible to calibrate a small flow rate of 1 sccm when the MFM output was completely unstable. Ultimately, we insulated the piping installed up to the FCS_® to provide stability, and connected a series of two MFM devices. We confirmed that there were no differences in the values of the two devices, and used different devices for the adjustments and for the inspections. So in total we used four MFM devices to check every flow rate for the FCS_®. In addition, the MFC is listed in the catalogue as having a precision of ±1% at full scale, and covers more than 50% of the calibration range from the Chemicals Evaluation and Research Institute, Japan, which means that the MFM can also be used in the same way to cover more than 50%. Also, the zero point changes when the MFM is switched off, so we created an automatic adjustment and inspection device for the FCS_® that can calibrate even with zero point fluctuations. Furthermore, the flow rate precision is affected by errors in the mV voltage of the signal generator, so it was calibrated with a voltage reference.

The steps required to calibrate the flow rate in the company were explained to the Chemicals Evaluation and Research Institute, Japan, and we worked towards improving the precision with the same steps.

The MFC has a flow rate precision of ±1% full scale. For example, a MFC with a flow rate of 100 sccm is ±1 sccm. A 10% flow rate set at 10 sccm also has a flow rate precision of ±1 sccm. Therefore, the flow rate precision at 10%, is ±10% of the set flow rate. The FCS_® guarantees a precision of ±1% set point. So at 100 sccm, it is ±1 sccm, and at a 10% setting of 10 sccm, it is ±0.1 sccm. This is ensured by correcting the zero point temperature drift of the pressure sensor and the pressure at every flow rate with a digital control circuit.

By matching the flow rate with the flow rate reference device with traceability from the National Metrology Institute of Japan, differences between each FCS_® device were removed. The semiconductor process has become complicated, and this answers the strong demands in the industry for an improvement in flow rate precision with a reduction in differences between each device.

3.1 Impact of the IGS_® on people's lives and manufacturing operations

By adopting the IGS_®, maintenance time (time to replace parts) has been drastically shortened. It is especially bad for special gases used in semiconductor manufaturing to touch the air. This is because it reacts with the moisture in the atmosphere, which causes corrosion and contamination in the metal and particles. Although replacements are done with inert gas flowing through during maintenance, it is not simply a case of shortening the time it touches the atmosphere, but also reduces the possibility of these problems occuring.

The IGS_® has also been popular with branch boxes in semiconductor plants. Branch boxes are used to branch and distribute gas in a large diameter pipe to multiple smaller diameter lines, and it is constructed with special gases and inert gas for purging. By planning an estimated expansion, the expansion can be done without stopping the gas. With semiconductor plants, the building is built entirely in the first period of construction, but there are many cases where the productivity of the facility is increased according to demand. It has become possible to expand without stopping all the gas.

3.2 Impact of the ECV_® on people's lives and manufacturing operations

In terms of the impact on manufacturing operations, by adopting the ECV_®, air supply systems such as compressors which generate the air required to power conventional pneumatic valves are not required, and can since it can be powered with just electricity, it is possible to save on manufacturing equipment.

3.3 Impact of the FCS_® on people's lives and manufacturing operations

With the general concept of the flow factor, inventory of spare parts in semiconductor plants has drastically decreased. With the MFC which is manufactured for every gas type and flow rate, unexpected situations are prepared for by storing an inventory of spare parts for every MFC's gas type and flow rate range close to the semiconductor plant (When a problem occurs, it is standard to deliver parts within 24 hours around the world). This costs an enormous amount of money. By adopting the flow factor in the FCS_®, it is possible support these situations by just storing one to two devices with the flow rate range in offices around the world.

Since the FCS_® is not affected by fluctuations in the gas supply pressure, pressure regulators and pressure guages for monitoring in semiconductor manufacturing equipment gas supply systems are not required. With this, it is possible to reduce the cost of these parts, and simultaneously make the gas panel much more compact. This makes it possbile to reduce the amount of space occupied in semiconductor plant clean rooms which are expensive to maintain and manage.

As a result of aiming for thorough quality control with traceability and the flow rate reference device, difference between each FCS_® device are extremely small. With the MFC, when the manufacturer changes and the flow rate shifts, or even when the same manufacturer replaces the MFC, a strict flow rate check is done, but by adopting the FCS_®, this is completely unnecessary.