JPH04155937A - Monitoring of overpower of laser beam - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Summary of the Invention] Regarding an overpower monitoring method for determining the intensity of laser light for cutting thin films such as fuses, the laser power range can be easily and quickly determined, and automation is also easy. In order to provide a power monitoring method, a second thin film layer is provided below the first thin film layer to be cut with a laser beam, with an insulating layer in between, and the electrical resistance of the second thin film layer is The upper limit of the intensity of the laser beam that cuts the first thin film layer is detected based on the change in the measurement value.

[Industrial application field]

The present invention relates to an overpower monitoring method for determining the intensity of laser light for cutting thin films such as fuses.

In recent memory ICs, switching to redundant circuits is achieved by cutting laser fuses, thereby significantly improving yield. Also, for ICs that require precise resistance values,
Resistance trimming is performed using laser light.

With such a laser beam, if the energy is too low, it will result in poor cutting, and if it is too large, there will be problems such as large holes reaching the underlying substrate, contaminating the substrate, and leakage current. must be within an appropriate range.

[Conventional technology]

The laser fuse is initially kept in a connected state and, if necessary, is cut by a laser beam to be in a disconnected state. To explain with Figure 5, (a) is the initial state, 10 is the laser fuse,
Surrounding it, 12 is an insulating layer (SiO□), 14 is a semiconductor substrate, and 16 is a diffusion layer formed on the substrate. (b) shows a state in which the fuse 1O is cut by laser beam (giant pulse) projection. Since the energy of this laser beam was appropriate, the fuse was cut with an appropriate hole 18 (completely cut with no uncut portion remaining), and there was no large hole. (C) is an example in which the energy was excessive, and although the fuse 10 was completely cut, a large hole 18a reaching the board 14 was left. In such a state, there are problems such as the melted fuse lO material contaminates the substrate 14 and a leakage current is generated between the fuse and the substrate or between the diffusion layer 16 and the substrate (pn junction).

As shown in FIG. 6(a), the fuse 1O consists of a wide part that becomes a terminal part and a narrow part in the middle, and cutting is performed by projecting a giant pulse P onto this narrow part. To give a numerical example, the diameter of giant pulse P is 6
.mu.m, and the width of the narrow part of the fuse is 2 .mu.m.

Conventionally, to optimize the power of the laser beam, the power is changed in appropriate steps, the fuse is cut with the power of each step, and the power that changes the fuse from a defective cut to a cut length is set to the minimum value, and a large hole is created. The method is to set the power to the maximum value and adopt an appropriate value between these values, but cutting fuses and generating large holes is done visually using an ordinary microscope or scanning electron microscope (SEM), which is quite complicated. It's a lot of work. For example, when examining a large hole by taking a cross section at that part, it is not easy to cut (separate the chip) at the center (diameter part) of the hole.

[Problem to be solved by the invention]

As described above, in the conventional laser power determination method, not only is the process cumbersome, but since it is based on visual inspection, there are individual differences, and accurate and stable power determination cannot be performed.

Also, it is impossible to automate power determination.

The minimum power can be easily determined by measuring the resistance value of the fuse after laser irradiation. That is, if it is broken, the resistance value is close to ω, and if it is not broken, it is a short circuit (actually a certain resistance value), so the minimum power can be determined using this and automation is easy.

However, the only way to determine the maximum power is to visually observe the shape of the fuse layer, the insulating film underneath it, the semiconductor substrate, etc. It is impossible for a surface unevenness measuring device to measure 1 μm or less, and even if it is a large hole, the depth is 1 μm or less, so a surface unevenness measuring device cannot measure it accurately. Also S
In EM, it is necessary to take a cross section, which takes a considerable amount of time (for example, one day) and effort, and cannot be automated.

SUMMARY OF THE INVENTION An object of the present invention is to improve such a problem and provide a power monitoring method that can easily and quickly determine the range of laser power and can be easily automated.

[Means to solve the problem]

As shown in FIG. 1, in the present invention, a second thin film layer 2 is placed below a first thin film layer (the above-mentioned fuse) 10 with an insulating layer in between.
Set 0. The thin film layer 20 is a diffusion layer formed on the semiconductor substrate 14 in FIG. 1, but as shown in FIG.
A conductive layer or a resistive layer formed on the semiconductor substrate 14 via an insulating layer may be used. 22a and 22b are lead wires connected to both ends of the second thin film layer 20. The fuse 10 that is actually used is attached with a lead wire or a terminal portion, but since the fuse used here is for testing (for determining laser power), neither the lead wire nor the terminal portion is necessary.

[Effect]

A plurality of such test fuse sets are formed on a chip, and the fuses are cut by projecting a laser beam (giant pulse) while changing the intensity at appropriate steps. and lead wire 22a.

The electrical resistance of the second thin film layer 20 is measured via 22b.

If a change occurs in the second thin film layer 20, it means that the laser beam was too strong and a large hole was formed. Therefore, the upper limit of the laser beam intensity is the laser beam intensity of the step immediately before this step, and the intensity of the laser beam used for cutting the fuse is set to be below this upper limit.

The lower limit of the laser beam intensity is the minimum value of the laser beam intensity that blows the fuse 10, and this can be detected by turning on/off the fuse 10 actually used. If a fuse with a terminal is also prepared on the test chip, the lower limit can be determined using this fuse, and the laser light intensity to be adopted can also be set to an intermediate value between the lower limit.

〔Example〕

The resistance of the second thin film layer 20 is 5Ω to 200Ω in sheet resistance.
, the total resistance is 500Ω to 2Ω. In order to increase the rate at which the cross-sectional area of a portion of the diffusion layer changes when it is ablated by a laser beam, it is desirable to form the diffusion layer into a bar shape with a width of 1 μm or less.

By doing this, even if only a very small surface is scraped, the rate of change in resistance value is large and detection becomes easy.

When the second thin film layer 20 is a diffusion layer formed on the substrate 14 as shown in FIG. 1, a change in this resistance value indicates a loss of the substrate, which should not occur. The intensity of the laser beam should be weaker than this. However, since resistance changes cannot be measured unless the second thin film layer is damaged, the intensity of the laser beam used to cut the thin film (fuse) must be set to a value sufficiently lower than the intensity at which resistance changes can be measured. Set to .

A specific procedure for measuring changes in resistance value is shown in FIGS. 2 and 3. In FIG. 2(a), two sets of fuses 10 with thin film layers 2o are prepared. Then, a laser beam is irradiated onto the negative side A, and no laser beam is irradiated onto the other side B. Since A and B are made in the same size and in the same process, they can be expected to have the same resistance value, and if the resistance value of A becomes greater than the resistance value of B, it can be determined that A has been damaged by laser irradiation.

Note that it is possible to detect resistance changes even with a single element by measuring resistance twice, before and after laser irradiation, but realistically, it is desirable to measure resistance only once, and Figure 2 shows this point. is valid. In this figure 2 (a), 24 (subscripts a, b,...
. . . are used to distinguish them from each other and are omitted as appropriate) are terminals (pads) for resistance measurement, which are connected to both ends of the thin film layer 20 by aluminum wiring.

FIG. 2(b) shows an example in which four fuses with thin film layers are provided and assembled into a bridge. Laser irradiation only for A, or A and D
When irradiated with laser light of the same intensity, it is possible to detect resistance value changes and the upper limit of laser light intensity with high sensitivity.

Figure 3 shows T e g (Test engineer
In order to increase the change in resistance value, 20 fuses with thin film layers were installed, and 10 fuses were connected in series between terminals 24b and 24a and between terminals 24b and 24c, and one group A was connected to the laser. The other group B was used for laser irradiation, and the other group B was used for non-laser irradiation. This is used in the form shown in Figure 2(a).

FIG. 4 shows an example in which the thin film layer 20 is made of a conductor such as aluminum or polycrystalline silicon or a resistor 2OA. The thickness of the insulating layer between the test or dummy fuse lO and the thin film layer 20 is made equal to the thickness of the insulating layer between the fuse and the semiconductor substrate that is actually used. The thickness of the insulating layer between the two may be arbitrary.

If the thin film layer 20 is a diffusion layer formed on a substrate as shown in FIG. 1, it will correspond to the lower layer (semiconductor substrate) of the actual fuse,
The behavior of excessive power heat conduction and energy absorption closely match reality, making it an accurate monitor. Note that irradiation with excessive power will create a large hole that reaches the substrate, but this is avoided in FIG. 4.

To give a numerical example of the laser power, the lower limit is 0.4 μJ and the upper limit is 0.6 μJ. Therefore, the intensity of the laser beam used for cutting the fuse is 0.5 μJ, which is the intermediate value.

〔Effect of the invention〕

As explained above, in the present invention, the upper limit of the laser power can be measured by measuring the resistance value of the thin film layer provided below the fuse.
When combined with the same lower limit measurement that can be detected by measuring the fuse resistance value, it is possible to automatically measure the laser power range and determine the optimal value of the laser power for fuse blowing without human intervention. As a result, cutting can always be performed with appropriate laser power, eliminating cutting defects and making it possible to save labor and achieve automation.

[Brief explanation of drawings]

Figure 1 is the principle of the present invention. Figures 2 to 4 are explanatory diagrams of Examples 1 to 3 of the present invention. Figure 5 is an explanatory diagram of fuse cutting. Figure 6 is an explanatory diagram of the fuse and laser beam. It is. In FIG. 1, 10 is the first thin film layer, 20 is the second thin film layer, 1
4 is a semiconductor substrate, and 12 is an insulating layer.

Claims (1)

[Claims] 1. A second thin film layer (20) is provided below the first thin film layer (10) to be cut with a laser beam, with an insulating layer in between, and the electricity of the second thin film layer is By measuring the resistance and its change,
A method for monitoring overpower of a laser beam, comprising detecting an upper limit of the intensity of the laser beam that cuts the first thin film layer. 2. The laser beam overpower monitoring method according to claim 1, wherein the second thin film layer is a diffusion layer formed on the surface of the semiconductor substrate. 3. The laser light overpower monitoring method according to claim 1, wherein the second thin film layer is a metal or resistor layer formed on the semiconductor substrate with an insulating layer interposed therebetween. 4. The laser beam according to any one of claims 1 to 3, wherein the electrical resistance of the second thin film layer is detected as the difference between the electrical resistances of those irradiated with the laser light and those that are not irradiated with the laser light. Overpower monitoring method.