Test method for wire bonding craters

Wire bonding crater testing: the invisible guardian of integrated circuit package reliability

In the sophisticated world of microelectronic packaging, bonding wires are like "nanobridges" connecting chips and external circuits, and their performance directly determines the operational accuracy and lifespan of electronic devices. From smartphone processors to satellite communication modules, from sensors in medical devices to control units in new energy vehicles, these metal wires with a diameter of only a few microns to tens of microns shoulder the key mission of transmitting electrical signals and ensuring circuit connectivity. However, in the bonding process, a hidden defect called "crater" is quietly threatening the reliability of integrated circuits - it is hidden under the solder ball, difficult to detect through conventional inspection, but may cause fatal failures such as leakage and breakdown short circuit in equipment operation. Therefore, crater testing, as the core means of detecting such hidden dangers, has become an indispensable link in the quality control of integrated circuit packaging.

1. Crater: a hidden quality hazard lurking under the welding ball

1. The formation and morphological characteristics of craters

Pits are microscopic defects that cause damage to the aluminum layer of the chip pad and the underlying silicon compound due to improper bonding parameters or imbalance in process control during the bonding process of integrated circuit packages. When the contact, bonding, or bonding power during bonding exceeds the tolerance threshold of the chip pad, the weld ball will hit the surface of the pad like a "miniature ram": in mild damage, the crater will appear crescent-shaped depression, destroying only the surface aluminum metal; In severe cases, a ring-shaped pit is formed, and even the underlying silicon layer is exposed, and obvious traces of silicon loss can be seen under the microscope. This damage does not appear instantaneously, but is like a "hidden injury" in a circuit that gradually worsens over the long operation of the equipment.

2. Concealed hazards of craters

Unlike defects that can be detected through visual inspection, such as deballization and false welding, the crater is hidden at the contact interface between the solder ball and the pad and must be exposed through destructive testing. Its hazards have significant latency: it may only appear as a slight leakage or a decrease in reverse breakdown voltage in the early stage, but if not dealt with in time, the leakage will gradually intensify as the power-on time is prolonged, eventually leading to a continuous drop in the reverse breakdown voltage and even a breakdown short circuit. This failure will accelerate under the action of temperature stress or electrical stress - an automotive electronics test data shows that a chip with a crater defect has an 8 times more probability of failure after 1,000 hours of operation at 150°C. More seriously, craters can weaken the bond strength and medium insulation, laying hidden dangers for sudden equipment failures.

2. Crater test: Unveiling the mystery of microscopic defects

1. Test principle: the "perspective eye" of chemical corrosion

Crater testing is essentially a destructive detection method that utilizes chemical corrosion, and its core principle is to dissolve the surface metal of the chip pad with the solder ball through a specific chemical solution, exposing the underlying structure to observe the damage. Aluminum, as the main material of pads, has unique chemical properties - it is easily soluble in strong bases such as potassium hydroxide (KOH) and sodium hydroxide (NaOH), as well as acid solutions such as phosphoric acid (H3PO4) and hydrochloric acid (HCl), but is insoluble in water, which provides a precise and controllable basis for corrosion processes.

Commonly used chemical reactions in the test are as follows:

Strong alkali system: 2Al + 2KOH + 2H2O = 2KAlO2 (potassium metaaluminate) + 3H2↑, which quickly dissolves the aluminum layer through a strong alkaline solution and releases hydrogen at the same time.

Medium-strong acid system: 2H3PO4 + 2Al = 2AlPO4 (aluminum phosphate) + 3H2↑, using the medium-strong acid of phosphoric acid to slowly corrode aluminum to avoid excessive damage to the silicon layer.

These reactions selectively remove the metal layer while preserving the silicon substrate, making crater defects hidden under the solder ball visible, like putting a "microscopic perspective eye" on the inspector.

2. The key value of the trial: preventing batch risks

The difference in the material of the bonding wire directly affects the probability of crater generation - the gold wire process is mature and stable, and the damage to the pad is small, so it is mostly used for high-reliability products. Copper and aluminum wires are more likely to cause PAD damage when bonded due to their high hardness. For example, when a wafer factory is in trial production of a new product, it is found that the bonding force of copper wire bonding exceeds the standard value by 10% through the crater test, and the crater incidence rate of subsequent batches is reduced from 5‰ to 0.1‰ after adjusting the parameters in time. This preventive inspection can effectively avoid batch quality problems and reduce cost losses due to rework.

3. Standardized process and safety specifications for crater testing

1. Preparation before the test

The safety of crater testing is of utmost importance, as highly corrosive chemical agents are involved and must be carried out in fume hoods, and operators are required to wear acid and alkali resistant overalls, protective masks, and nitrile gloves. Pretreatment of test samples includes:

Exposed chip surface: The packaged product needs to be opened first to ensure that the pad area is completely exposed.

Pick off the welding wire: Use a special pick-up needle to cut off all the welding wires from the second soldering point (fishtail) to avoid secondary damage to the pad caused by the force of the wire arc in subsequent operations.

These steps may seem simple, but they directly affect the accuracy of test results - a laboratory once caused 30% of samples to be artificially damaged during chip removal due to failure to completely fault the solder wire, misidentifying it as a crater defect.

2. Refined operation of corrosion and detection

The core process of the trial consists of four key links:

Solution configuration: Reagents are selected according to the material of the pad, aluminum pads are often used with 5%-10% KOH solution or 85% phosphoric acid solution, and nickel-palladium gold pads need to use aqua regia (concentrated hydrochloric acid: concentrated nitric acid = 3:1);

Heating corrosion: Place the beaker containing the solution on the heating plate, control the temperature at 60-90°C (the temperature of the strong alkali system is lower, and the strong acid system can be appropriately increased), and continue to observe after putting it in the sample until the solder ball is completely detached - this process needs to avoid excessive corrosion, otherwise it will destroy the morphology of the silicon layer;

Cleaning and drying: Put the corroded samples into a deionized water ultrasonic cleaning machine, wash them at a frequency of 40kHz for 3-5 minutes to remove residual reagents, and then use dust-free filter paper to absorb the surface moisture.

Microscopic Observation: The surface of the Pad is examined using a 200-500x metallurgical microscope, recording crater features such as craters and silicon layer exposure.

Practical data from a semiconductor packaging factory shows that strict adherence to the process can achieve test repeatability of more than 95%, greatly reducing the false positive rate.

4. Comparison of test methods of different pad processes

The differences in the process of chip pads determine the choice of test methods, and the three mainstream types of pad processes currently have their own suitable testing solutions:

1. 镍钯金焊盘（NiPdAu Pad）

The surface of this type of Pad is a gold layer and needs to be corroded with aqua regia - the strong oxidation properties of aqua regia quickly dissolve the gold layer (reaction formula: Au + HNO3 + 4HCl = H [AuCl4] + NO↑ + 2H2O), exposing the underlying nickel-palladium structure. Its advantage is its fast corrosion rate (usually completed in 3-5 minutes), but its disadvantages are also obvious: aqua regia is extremely corrosive to experimental equipment, and the residual palladium-nickel compounds are difficult to completely remove, which may interfere with observation. Therefore, the method requires a dedicated corrosion-resistant vessel and an additional nitric acid immersion step after corrosion to remove residue.

2. 铝焊盘(Al Pad)与铜铝焊盘(CuAl Pad)

Due to the aluminum layer on the surface, you can flexibly choose a strong alkali or strong acid system:

Sodium hydroxide / potassium hydroxide solution: suitable for rapid detection, can corrode the aluminum layer at room temperature, complete the reaction in 10-15 minutes, and the surface treatment is clean, and the damage to the silicon layer is small; However, the concentration needs to be strictly controlled (5%-8% recommended), and high concentrations of lye will lead to silicon layer etching (reaction formula: Si + 2NaOH + H2O = Na2SiO3 + 2H2↑), resulting in false positive results.

Phosphoric acid/hydrochloric acid solution: Preserves the Pad morphology through the passivation characteristics of aluminum (an oxide film forms on the surface to prevent further reactions), suitable for observing the details of aluminum layer damage; However, the corrosion time is longer (20-30 minutes), and the temperature (60-70°C) needs to be precisely controlled, otherwise the efficiency will be affected by the slow reaction.

In practical applications, strong alkali systems account for more than 70% of aluminum pad detection due to their strong stability and easy operation. To improve accuracy, acid-base comparison tests can be used to reduce the false positive rate to less than 0.5% by treating the same batch of samples with KOH and phosphoric acid, respectively, and if the results are consistent, they are reliable.

5. The impact of craters on reliability: structural design determines the risk level

The degree of harm of craters is not generalized, but is determined by the internal structure design of the chip pad. Through the comparative test of the two chips, CUP (with circuit under pad) and NO CUP (without circuit under pad), it is clear that the following is true:

CUP chip: Active circuitry is distributed under the pad, and crater damage can directly destroy the circuit connection, and after 3 times of 260°C reflow soldering, 500 temperature cycles (-55°C to 125°C), or 96 hours of high-pressure cooking (121°C, 100% RH), the probability of failure of the functional test reaches 35%;

NO CUP chip: The bottom layer of the Pad is a same-potential metal layer, even if there is a crater, it will not affect the electrical performance, and the functional test pass rate after the above reliability test is still maintained at 100%.

This difference stems from the structural design - the Top1 and Top2 layers of the NO CUP chip are equipotential, and the damage to the first layer does not affect the overall circuit. There is a potential difference between the Top2 and Top3 layers of the CUP chip, and any damage may lead to leakage or short circuit. Therefore, the chip design stage should optimize the Pad structure, such as controlling the thickness of the aluminum layer above 0.8-1μm (too thin can easily form cavities and increase the risk of craters), and setting an insulating layer under the pad.

6. Conclusion: The gatekeeper of quality in technological evolution

As the "last line of defense" for integrated circuit packaging quality control, crater testing is advancing simultaneously with chip process innovation. With the popularization of advanced technologies such as 3D packaging and chiplets, the Pad structure is becoming more and more complex, which puts forward higher requirements for test methods - for example, for the hybrid Pad structure of heterogeneous integrated chips, it is necessary to develop a gradient concentration corrosion solution to achieve selective dissolution; In the face of nanoscale pad size, it is necessary to combine scanning electron microscopy (SEM) and energy spectroscopy (EDS) to improve detection accuracy.

For packaging technicians, mastering the essence of crater testing lies not only in proficiency in the operation process but also in understanding the relationship between material properties and structural design – from chemical solution ratio to precise temperature control, from microscopic topography identification to risk level assessment, every step embodies the ultimate pursuit of reliability. Today, as electronic devices move towards high integration and high reliability, crater testing is guarding the safe operation of each chip with its unique "perspective" ability and becoming an invisible guardian of the high-quality development of the integrated circuit industry.