How to verify the wafer processing equipment EFU project? Prototype, first unit testing and batch confirmation

EFU verification of wafer processing equipment is not the same thing as purchasing a general-purpose FFU and installing it on the ceiling. After the equipment-side EFU is installed into the wafer processing equipment, it faces the back pressure inside the equipment, the air flow disturbance caused by the movement of the robotic arm, and the sensitive requirements of the process itself on particles. The qualified rated parameters of the EFU unit does not mean that the target wind speed distribution can be maintained after being installed in the equipment; the data in the prototype stage is beautiful, but it does not mean that it can be reproduced in each batch. Verification is divided into three steps: the prototype stage to confirm the design feasibility, the first stage to confirm the matching of the whole machine, and the batch stage to confirm the consistency . The test items, record content and judgment basis of the three stages need to be agreed separately. Prototype data cannot be used to replace the first unit verification, nor can the first unit data be used to cover batch sampling inspections. Aili Industrial's 70mm ultra-thin EFU is organized into prototypes and first units for verification based on confirmation drawings in the wafer processing equipment project, and air volume, wind speed distribution and vibration data are archived by stage.

Prototype stage: first confirm whether the design direction is correct

The goal of the prototype stage is not to 'prove that the EFU can meet the standards', but to confirm whether the air duct structure, fan selection and filter configuration are feasible under the equipment installation conditions . At this stage, imperfect data is allowed, but enough information needs to be obtained to determine what to change next.

The prototype test items should cover the core performance indicators of the device-side EFU. The air volume and static pressure should be measured under the actual back pressure conditions of the equipment, rather than on an open bench - the air duct turns, filter resistance and air outlet obstruction inside the equipment will cause the fan working point to deviate from the sample curve. The wind speed distribution on the air outlet surface needs to be measured at the target installation location. The measurement points are divided according to the air outlet surface grid, and the wind speed and deviation range of each point are recorded to determine whether the laminar flow meets the coverage requirements of the wafer path. The temperature rise in continuous operation should be simulated according to the actual inlet air temperature inside the equipment, and the temperature rise curve from cold start to thermal stability should be recorded, rather than just measuring one time point. Vibration needs to record the rated speed and amplitude within the commonly used speed adjustment range. Wafer processing equipment is sensitive to micro-vibration, and the vibration transmission path of the EFU must be exposed during the prototype stage.

The particle removal performance of EFU on the semiconductor device side has a special test method reference. The SEMI F111 standard describes the testing of particle removal capabilities of equipment fan filter units (EFFU) installed inside equipment, emphasizing the need to consider filter media specifications, air flow and integrity of the EFFU components together, rather than looking at filter efficiency alone. Particle testing at the prototype stage can refer to this idea and evaluate the system performance of the EFU in the simulated equipment cavity or actual installation location, rather than relying solely on the filter factory certificate.

The prototype data and confirmation drawings are archived together as the basis for modifying the configuration of the first unit. If it is found during the prototype stage that the wind speed distribution does not meet the standard or the temperature rise exceeds expectations, the air duct structure or fan specifications should be adjusted before the first unit, rather than bringing the problem to the first unit.

First stage: Confirm the matching of the whole machine and the performance of the device

The first verification is a complete test of after the design direction of the prototype is confirmed the first EFU made according to the final configuration . The difference between the first unit and the prototype is that the prototype may allow temporary changes and repeated debugging, while the first unit should represent the delivery status of subsequent batches.

The items tested on the first machine are more complete than those on the prototype, and the judgment criteria are clearer. In addition to air volume, wind speed distribution, temperature rise and vibration, it is also necessary to add interlock verification between the EFU and the equipment main control : whether the startup command is executed according to the process rhythm, whether the speed adjustment command is consistent with the set value, whether the fault alarm is correctly read by the equipment PLC, and whether the controller's action is as expected when communication is interrupted. For EFU groups controlled by partitions, the first stage should check the correspondence between the slave address and the physical location one by one to confirm that the start and stop of the partition will not accidentally trigger adjacent areas.

Another focus of the first machine verification is the maintenance path . After the EFU is installed in the wafer processing equipment, it is not easy to judge from the drawings whether there is enough space for filter replacement and fan maintenance, whether other parts need to be disassembled, and whether the sealing state can be restored after replacement. It can only be confirmed by an actual operation after the first unit is assembled. For the method of checking the filter replacement direction and space constraints, please refer toHow to choose EFU filter without partition? Efficiency, resistance, sealing and complete machine matching methods are explained.

The test data of the first unit is compared with the prototype data to confirm whether the design modification is effective. At the same time, the final configuration of the first unit (fan model, filter specifications, control parameters, and installation interfaces) is solidified into the equipment BOM. After the first one passed, this configuration became the benchmark for mass production.

Batch confirmation: Confirm that each unit is the same as the first one

The verification goal in the batch phase has changed from 'whether the performance meets the standards' to 'whether the consistency is controllable'. No matter how good the data of the first unit is, if the air volume, wind speed or vibration dispersion of the batch EFU is too large, the cleanliness of the equipment will still fluctuate.

Batch confirmation usually adopts the method of first-piece full inspection plus sampling inspection . The first piece is tested one by one according to the standards of the first piece to confirm that the batch process is consistent with the first piece. Subsequent batches will be randomly inspected according to the agreed proportion. The random inspection items can focus on the indicators that have the most direct impact on the equipment : the air volume at the target working point, the uniformity of wind speed on the air outlet surface, the rated speed vibration value, and the address and speed regulation mapping of control communication. When it is found during the random inspection that a certain indicator deviates from the first benchmark and exceeds the allowable range, the random inspection ratio should be expanded or returned to full inspection.

The judgment of batch confirmation is based on the benchmark data archived by the first unit, not the rated value in the catalogue. The air volume of the first unit may be 600 m³/h under the back pressure of the equipment. For batch sampling, the allowable deviation is set around 600, instead of taking 800 on the sample as the qualification line. The random inspection records are compared with the data of the first unit, batches with deviations within the agreed range are released, and batches with deviations out of tolerance are determined after the reasons are found out. Aili Industrial organizes production according to the confirmed configuration of the first unit during the batch supply of EFU. The random inspection data is compared with the first unit's benchmark, and the control parameters and installation interfaces correspond one by one according to the confirmation diagram.

Misunderstandings to avoid

It should be avoided to use the prototype data to directly release the first unit. The prototype is allowed for debugging but the first unit represents the batch status; it should be avoided that the first unit only measures the EFU monomer without verifying the interlocking with the equipment main control; it should be avoided that the batch inspection uses the sample book rating as the qualification line without looking at the actual benchmark of the first unit under the equipment back pressure; it should be avoided that the random inspection only measures the air volume but not the wind speed distribution and vibration. The impact of these three items on the wafer processing equipment is equally direct. The core of the EFU project verification is not 'a certain machine has been tested', but the direction of the prototype, the benchmark of the first machine, and the benchmark of the batch. The data in the three stages form a traceable chain.

Related questions and answers

Can the prototype and the first machine be merged into one stage?
Not recommended. The purpose of the prototypes is to expose design issues and allow for modifications, with the first unit representing the final delivery state. If merged, temporary changes in the prototype stage may be carried into the batch, or the first unit may not have baseline significance because it 'has to be changed'. When the equipment structure is relatively simple and the EFU configuration is close to the standard model, the interval between the two stages can be shortened, but the recording needs to be separated.

Does the first verification have to run the entire process cycle?
At least the typical operating states of the EFU in the equipment need to be covered: early start, speed switching, delayed shutdown, and communication interruption simulation. The complete process verification cycle of wafer processing equipment is long. The first verification of EFU can focus on the response and linkage of the EFU itself, without waiting for the completion of process verification of the whole machine. However, the confirmation of the linkage logic must be completed at the first stage and cannot be left until the batch.

How many units will be sampled for batch inspection?
Depends on the batch size and the dispersion of the first data. For small batches (for example, less than 10 units), the first unit can be fully inspected and 1-2 units can be randomly selected from the rest; for larger batches, 5%-10% of the units can be randomly inspected, but not less than 2 units. If an indicator is found to be close to the limit of the allowable deviation during the random inspection, the random inspection should be expanded. The specific proportion can be agreed in the procurement technology agreement.

Related articles

How to configure advanced packaging equipment EFU? Key points of clean area, air volume filtration and equipment interface

How to choose EFU filter without partition? Efficiency, resistance, sealing and complete machine matching methods

How to choose ultra-thin EFU fan? Space, air volume, static pressure, heat dissipation and maintenance points

Key points for selection and acceptance inspection

Aili Industrial EFU selection and procurement support

Aili Industrial's 70mm ultra-thin EFU is organized and verified in three stages: prototype, first unit and batch in the wafer processing equipment project. In the prototype stage, the air volume and wind speed distribution are tested according to the equipment installation conditions. In the first unit stage, the configuration is solidified and the interlocking with the equipment main control is verified. The batch inspection is based on the first unit benchmark. The air volume, temperature rise, and vibration data are archived together with the confirmation chart, making it easier for the equipment factory to trace the correspondence between each batch.

Learn about Aili Industrial EFU products and customized configurations. Check out the compact models.70mm ultra-thin EFU , the entire machine structure can be viewedEFU equipment fan filter unit , the control scheme can be viewedDC speed regulating EFU equipment fan filter unit , non-standard interface can be viewedCustom EFU . Please submit the equipment installation conditions, target wind speed and expected batch size to facilitate confirmation of prototype test items and first unit verification plan.