CN107358977A - A kind of test method that single-particle soft error is carried out with X ray - Google Patents

Abstract

The invention discloses a test method for single particle soft error by using X-rays, which comprises the following steps: loading a bias voltage lower than the normal voltage on an SRAM test device; irradiating the SRAM test device with X-rays, and recording test data; and Calculate the corresponding single event turnover cross section under the bias voltage according to the experimental data. The present invention confirms that the secondary electrons generated by nano-scale devices in the process of X-ray irradiation can cause the device to flip by using X-ray single-particle soft error test method. For the study of single event effects of devices, a ground simulation source is added. Compared with other radiation sources, it has the advantages of accurate dose rate control, easy radiation shielding, no radiation residue, and easy access to machine time.

Description

An experimental method for single event soft errors using X-rays

technical field

The invention belongs to the technical field of space radiation effect and reinforcement. More specifically, the present invention relates to a method for testing single event soft errors using X-rays.

Background technique

Static random access memory (SRAM) is widely used due to its advantages of fast read and write speed, low power consumption, and high integration. As one of the essential memories in spacecraft and satellites, it plays a huge role in data storage.

For satellites that need to perform space missions, in the outer space environment, there are a large number of galactic cosmic rays, solar flares, and particles in the geomagnetic capture belt, which will have a non-negligible impact on the reliability of electronic equipment on the satellite. The lifespan of the satellite. Among many radiation effects, the single event reversal effect has attracted much attention as the main radiation effect. Single event upset (Single Event Upset) means that when a particle passes through the chip, energy will be lost along its path, and a large number of hole-electron pairs will be generated. These charges are collected through drift and diffusion, and a transient pulse is generated to make the logic state of the device A rollover occurs.

With the continuous development of integrated circuits to high integration and low feature size, the gate length, node depth, and oxide layer thickness of devices are all reduced accordingly. Figure 1 shows the critical charge required for device flipping under different feature sizes. Judging whether a single-event upset occurs in a device depends on the magnitude of the critical charge (Q _crit ). The critical charge is defined as the minimum charge collected by the sensitive electrode that can cause the logic state of the device to flip. It can be seen that the smaller the feature size of the device, the smaller the value of the critical charge, and the easier it is for single event flipping to occur.

The simulation sources usually used to simulate single event effects experiments are heavy ion accelerators, pulsed lasers, transuranium nuclides ( ²⁵² Cf) and so on. Through the above-mentioned simulation sources, a large number of studies on single event effects have been carried out on SRAM devices at home and abroad. As a kind of radiation source, X-ray is mostly used in the study of total dose effect on SRAM devices due to its particularity, and there are few reports on the study of single event effect. There are many ways for X-rays to interact with matter. At low energy, photons mainly react with extranuclear electrons. The types of reactions include photoelectric effect, Compton scattering, and electron pair effect. In the study of the single event effect, the main consideration is the ionization damage caused by the photoelectric effect that occurs at the energy of 30keV-100keV to the device. The multi-layer metal wiring layer in the device, especially the high atomic number material (such as W) contained in the metal wiring layer and the sensitive area of the device (mainly composed of SiO ₂ ) constitute an interface with a large difference in atomic number, and dose enhancement occurs. Effect, a large number of secondary electrons are generated, and these secondary electrons will cause additional energy deposition inside the device. The energy deposition of these secondary electrons is much larger than the energy deposition caused by the photoelectron itself, and the additional energy is deposited in the sensitive area of the device. , which will greatly increase the probability of single-event upset of the device.

For nanoscale devices, especially when the characteristic size of SRAM devices reaches 40nm and 28nm, the single event sensitivity of the device increases, and the single event flipping caused by secondary electrons generated by the photoelectric effect will become more and more significant. Obtaining the X-ray single event upset cross-section through experiments provides an important technology for evaluating the ability of devices to withstand single event upset.

Contents of the invention

An object of the present invention is to solve at least one of the above-mentioned problems or disadvantages and to provide at least one advantage described hereinafter.

Still another object of the present invention is to provide a test method for single event soft errors with X-rays, which can obtain corresponding single-particle flip cross-sections under different bias voltage conditions, and confirms that nano-scale devices can withstand X-ray irradiation. During the process, the secondary electrons generated can cause the device to flip. Compared with previous studies, a ground simulation source has been added to the study of single event effects of nanoscale devices.

In order to realize these purposes and other advantages according to the present invention, a kind of test method of carrying out single event soft error with X-ray is provided, comprises the steps:

Apply a bias voltage lower than the normal voltage to the SRAM test device;

Perform X-ray radiation on the SRAM test device and record the test data; and

Calculate the corresponding single event turnover cross section under the bias voltage according to the experimental data.

Preferably, wherein, the following steps are included:

Step 1, pre-testing the SRAM test device;

Step 2. Arranging the X-ray optical machine;

Step 3, applying different bias voltages lower than the normal voltage to the SRAM test device;

Step 4: Carry out X-ray radiation to the SRAM test device, record the test data, and calculate the corresponding single event turnover cross section under different bias voltages according to the test data.

Preferably, wherein, in the step 1, it specifically includes: uncapping the SRAM test device to expose the chip, and after the uncapping is completed, the chip is powered on for a test to ensure that the chip performs normal data reading and writing, and Group and number the qualified SRAM test devices.

Preferably, wherein, the second step specifically includes: fixing the irradiation plate on the xyz precision mobile platform through a fixture, and using a laser collimator to ensure the center of the target chamber of the X-ray optical machine and the exposure of the SRAM test device parts on the same level.

Preferably, wherein, the step 4 specifically includes: before starting the X-ray radiation, writing the initial data 55H, and then adjusting the voltage to different bias voltages in the step 3, performing X-rays on the SRAM test device Radiation, after the X-ray radiation is over, read back the data, record the number N of single-event flips, and the flux Φ, and calculate the single-event flip cross-section σ of the X-ray irradiated SRAM test device under different bias voltages ,Calculated as follows:

Wherein, in the formula, M is the total number of bits of the SRAM test device, and A _cell is the chip area of the SRAM test device, and the unit is cm ² .

Preferably, wherein, the step three specifically includes: respectively applying 0.5V, 0.7V, 0.9V, or 1.1V to the SRAM test device with different bias voltages lower than the normal voltage of 1.5V, and loading 50kV tube voltage and a tube current of 30mA.

Preferably, it also includes the step of placing a 1mm thick Al plate filter in front of the SRAM test device, in order to filter out the low-energy part of X-rays and prevent the total dose effect from affecting the test results.

Preferably, wherein, after the end of the X-ray radiation, the read-back data specifically includes: after the end of the X-ray radiation, adjust the bias voltage to a normal voltage state, and read back all the data, and compare with the initially filled data. In contrast, if an error occurs, record the wrong data and the wrong address; then fill the data again, and read back the data to ensure that no errors occur during the read back process, and the SRAM test device can perform normal data reading and writing. The purpose is to ensure that the flipping error of the device during irradiation is a single-event soft error caused by high-energy electrons, rather than an error caused by the cumulative total dose effect.

The present invention at least includes the following beneficial effects:

1. In the present invention, it is proposed that under different bias voltage conditions, the corresponding single-particle flip cross section can be obtained, and it is confirmed that the secondary electrons generated in the nanoscale device during X-ray irradiation can cause the device to flip. Compared with previous studies, a ground simulation source has been added for the study of single event effects of nanoscale devices;

2. The test method used in the present invention is different from the method adopted by other radiation sources for single particle tests. By this method, a lower bias voltage can be loaded in the irradiation test, and a smaller critical mass can be obtained. The charge value makes it easier for the device to undergo a single-event flip. With the development of technology, the device's operating voltage will gradually decrease due to power consumption considerations. Through this method, the irradiation research of devices at lower voltages can be carried out in advance;

3. Compared with other radiation sources, the X-rays used in the present invention are valuable in heavy ion accelerators, and there are residues after irradiation in proton accelerators. The penetration depth of pulsed lasers in opaque media is not as good as that of ordinary light. X-rays have Accurate dose rate control, easy radiation shielding, no radiation residue, easy access to machine time, etc.

Other advantages, objectives and features of the present invention will partly be embodied through the following descriptions, and partly will be understood by those skilled in the art through the study and practice of the present invention.

Description of drawings

Figure 1 is a schematic diagram of the critical charge at different feature sizes;

Figure 2 is a schematic diagram of a standard 6-tube storage unit;

Fig. 3 is the schematic diagram of loading bias voltage in radiation test in one embodiment of the present invention;

Fig. 4 is a schematic diagram of the operation flow in one embodiment of the present invention;

FIG. 5 is a schematic diagram of the results of single event upset after testing the SRAM device in one embodiment of the present invention.

detailed description

The present invention will be further described in detail below in conjunction with the accompanying drawings, so that those skilled in the art can implement it with reference to the description.

Technical principle of the present invention:

When X-rays irradiate the device, energy will be lost along its path, and a large number of hole-electron pairs will be generated. These charges are collected through drift and diffusion, and transient pulses are generated to change the data stored in the device.

The SRAM device is composed of an array of static volatile memory cells and its peripheral circuits. Its address decoding is located inside the SRAM, so it can read and write any storage unit. As shown in Figure 2, it is a standard 6-tube storage unit consisting of 4 NMOS tubes and 2 PMOS tubes. The four transistors N1, N2, P1, and P2 are connected end to end to form two inverters, forming a closed feedback loop and forming a bistable structure, which can store a bit of data "1" or "0". The two transistors N3 and N4 are used to control whether to read or write data.

When particles enter the interior of the transistor, especially the space charge region of the reverse-biased PN junction in the drain regions of N1 and P2 transistors is the sensitive region for single-event inversion of the device. If the particle is incident on the drain of the N1 tube, it is considered that the stored data is "1" at this time, which will cause a transient current inside the device, and the drain voltage of the MOS tube will drop from high level to low level, while the P1 tube continues to Conduction will cause the data in the storage unit to fluctuate and cannot maintain a normal state. At this time, the voltage at point Q drops, affecting the gates of N2 and P2, N2 tube is turned off, P2 tube is turned on, the node Q_ voltage rises, and at the same time, N1 tube is turned on, P1 tube is turned off, and through a series of feedbacks, the storage unit The "1" becomes "0".

The condition for the flipping of the device is that the total amount of charges collected is Q>Q _crit , otherwise, flipping does not occur. It can be seen from Figure 2 that the smaller the feature size of the device, the smaller the value of _Qcrit , and the easier it is for flipping to occur. For a device, the value of the critical charge is affected by multiple parameters. In formula (1), the dielectric constant of SiO ₂ (ε _SiO2 ) and the dielectric constant of air (ε ₀ ) are fixed values, and the cell area (A _cell ) and oxide layer thickness (t _OX ) are determined by the device itself, the only variable is the voltage (V _DD ), at different voltages, the value of the critical charge will vary

different. Through verification, when the voltage loaded on the device is lower than the normal voltage, it can still ensure that the device can work normally. Therefore, in the X-ray irradiation test, the method shown in Figure 3 can be used to obtain the lowest bias voltage of the device during the irradiation process through multiple experiments. In the state of this bias voltage , the critical charge value of the device will reach the lowest value, and the possibility of flipping is greatly increased.

Fig. 4 shows a schematic diagram of the operation flow in one embodiment of the present invention, which specifically includes the following steps:

Step 1. Open the cover of the SRAM test device to expose the chip. After the cover is opened, the chip is powered on for a test to ensure that the chip performs normal data reading and writing, and the SRAM test devices that pass the test are grouped and numbered;

Step 2. Fix the irradiation plate on the xyz precision mobile platform through the fixture, and use the laser collimator to ensure that the center of the target room of the X-ray machine and the exposed part of the SRAM test device are on the same horizontal line;

Step 3. Apply 0.5V, 0.7V, 0.9V, or 1.1V to the SRAM test device with different bias voltages lower than the normal voltage of 1.5V, and apply a tube voltage of 50kV and a tube current of 30mA;

Step 4. Before starting the X-ray radiation, write the initial data 55H, then adjust the voltage to the different bias voltages in the step 3, and perform X-ray radiation on the SRAM test device. After the X-ray radiation ends, set the bias voltage to Adjust to the normal voltage state, and read back all the data, and compare it with the initially filled data. If an error occurs, record the wrong data and the wrong address; then fill the data again, and read back the data to ensure During the readback process, no error occurred, and the SRAM test device can perform normal data reading and writing. The purpose is to ensure that the flip error of the device during irradiation is a single event soft error caused by high-energy electrons, rather than cumulative The error caused by the total dose effect of the X-ray irradiation SRAM test device under different bias voltages can be calculated by recording the number N of single event flipping and the flux Φ, and the calculation formula is as follows :

Wherein, in the formula, M is the total number of bits of the SRAM test device, and A _cell is the chip area of the SRAM test device, and the unit is cm ² .

In another embodiment, the present invention uses X-rays to perform a single-event soft error test method specifically includes the following steps:

Before the irradiation test, the SRAM test device needs to be uncapped to expose the chip. After opening the cover, it is necessary to conduct a power-on test on the uncapped SRAM test device to ensure that the SRAM test device can store data, remove unqualified devices at the same time, and group and number the SRAM test devices that pass the test. Fix the irradiated panel on the xyz precision mobile platform through the fixture. Use a laser collimator for collimation to ensure that the center of the target chamber and the exposed part of the device are on the same horizontal line;

Before the irradiation test starts, write the initial data 55H into the SRAM test device under normal voltage, and read back the data to make sure that the SRAM test device does not have any errors during the read-back process, and then stop Read back the data, keep the filled data no longer change, and adjust the voltage so that the adjusted bias voltage is lower than the normal power supply voltage of the SRAM test device, then start the DC X-ray machine, select the tube voltage of 50kV and the tube voltage of 30mA The current is irradiated, and a 1mm thick Al plate is placed in front of the SRAM test device, the purpose is to filter out the low-energy part of the X-rays and prevent the total dose effect from affecting the test results;

After the irradiation is over, adjust the bias voltage to a normal voltage state, and read back all the data, and compare it with the initially filled data. If an error occurs, record the wrong data and the wrong address. Then fill the data again, and read back the data to ensure that no errors occur during the read back process, and the device can perform normal data reading and writing. The purpose is to ensure that the flip error of the device during irradiation is a single-event soft error caused by high-energy electrons, rather than an error caused by the cumulative total dose effect;

By repeating the above test steps and adjusting different bias voltages, the corresponding single event flip cross sections under different voltages can be obtained.

According to the above method, the results of the single event inversion cross section of the ISSI61WV device under X-ray are shown in Table 1, and the inversion cross section and bias voltage curve are shown in Figure 5. It can be seen that when the loaded bias voltage is higher than 1.0V, the flip phenomenon does not occur, and only when the loaded bias voltage is lower than 1.0V, the device flips. In the case of fixed X-ray energy, it can be known from formula (1) that the lower the applied bias voltage is, the smaller the value of the critical charge is, and the easier it is for single event reversal to occur. When the bias voltage is higher than 1.0V, the secondary The energy deposited by the electrons is not enough to cause the device to flip, indicating that X-rays have a strong dependence on the bias voltage when single-event flip occurs.

Table 1 ISSI61WV device X-ray single event flip section experiment results

Bias voltage (V) Flip Section (cm ² ) 1.2 no rollover 1.1 no rollover 1.0 3.79×10 ^-14 0.9 4.74×10 ^-13

The technical problem solved by the invention is: under the premise of not changing the layout of the device and the steps of the production process, a test method for the single particle soft error of the device is proposed by using X-rays.

The number of devices and processing scales described here are used to simplify the description of the present invention. Applications, modifications and variations to the inventive method of testing single event soft errors with X-rays will be apparent to those skilled in the art.

As mentioned above, according to the present invention, under different bias voltage conditions, the corresponding single-particle inversion cross section can be obtained, which confirms that the secondary electrons generated in the nanoscale device during X-ray irradiation can cause the device to invert. Compared with previous studies, a ground simulation source has been added to the study of single event effects of nanoscale devices. The test method used in the present invention is different from the method adopted by other radiation sources for single particle tests. By this method, a lower bias voltage can be loaded in the irradiation test, and a smaller critical charge value can be obtained. , making it easier for the device to undergo single-event flipping. With the development of technology, due to the consideration of power consumption, the operating voltage of the device will be gradually reduced. In this way, the irradiation research of devices at lower voltages can be carried out in advance. Compared with other radiation sources, the X-rays used in the present invention are expensive in heavy ion accelerators, and there are residues after irradiation in proton accelerators. The penetration depth of pulsed lasers in opaque media is not as good as that of ordinary light, and X-rays have a dose rate Accurate control, easy shielding of radiation, no radiation residue, easy access to machine time, etc.

Although embodiments of the present invention have been disclosed above, it is not limited to the applications set forth in the specification and examples. It can be fully applied to various fields suitable for the present invention. Additional modifications can readily be made by those skilled in the art. Therefore, the invention should not be limited to the specific details and examples shown and described herein, without departing from the general concept defined by the claims and their equivalents.

Claims (8)

1. A test method for carrying out single particle soft error with X-ray, is characterized in that, comprises the steps: Apply a bias voltage lower than the normal voltage to the SRAM test device; Perform X-ray radiation on the SRAM test device and record the test data; and Calculate the corresponding single event turnover cross section under the bias voltage according to the experimental data. 2. the test method of carrying out single particle soft error with X-ray as claimed in claim 1, is characterized in that, comprises the steps: Step 1, pre-testing the SRAM test device; Step 2. Arranging the X-ray optical machine; Step 3, applying different bias voltages lower than the normal voltage to the SRAM test device; Step 4: Carry out X-ray radiation to the SRAM test device, record the test data, and calculate the corresponding single event turnover cross section under different bias voltages according to the test data. 3. the test method for carrying out single event soft error with X-ray as claimed in claim 2, is characterized in that, in described step 1, specifically comprises: SRAM test device is carried out uncap processing, makes chip bare, and uncap is completed Finally, conduct a power-on test on the chip to ensure that the chip can perform normal data reading and writing, and group and number the SRAM test devices that pass the test. 4. The test method for single particle soft error with X-rays as claimed in claim 2, characterized in that, in said step 2, specifically comprising: fixing the irradiated plate on the xyz precision mobile platform with a clamp, using The laser collimator ensures that the center of the target chamber of the X-ray machine and the exposed part of the SRAM test device are on the same level. 5. The method for testing single particle soft errors with X-rays as claimed in claim 2, wherein said step 4 specifically includes: before X-rays start to radiate, write initial data 55H, and then adjust the voltage To the different bias voltages in step 3, perform X-ray radiation on the SRAM test device. After the X-ray radiation is over, read back the data, record the number N and flux Φ of single event flips, and calculate them at different bias voltages. The single event turnover cross section σ of the X-ray irradiated SRAM test device under the set voltage is calculated as follows: <mrow><mi>&amp;sigma;</mi><mo>=</mo><mfrac><mi>N</mi><mi>M</mi></mfrac><mfrac><mn>1</mn><mrow><msub><mi>&amp;Phi;A</mi><mrow><mi>c</mi><mi>e</mi><mi>l</mi><mi>l</mi></mrow></msub></mrow></mfrac></mrow> Wherein, in the formula, M is the total number of bits of the SRAM test device, and A _cell is the chip area of the SRAM test device, and the unit is cm ² . 6. The test method for carrying out single event soft errors with X-rays as claimed in claim 2, characterized in that, in said step 3, specifically comprising: respectively loading 0.5V, 0.7V, 0.9V to the SRAM test device, or 1.1V different bias voltage 1.5V lower than the normal voltage, and loaded with a tube voltage of 50kV and a tube current of 30mA. 7. the test method of carrying out single particle soft error with X-ray as claimed in claim 2, it is characterized in that, also comprises the step: place the Al plate optical filter of 1mm thickness before SRAM test device, purpose is in order to filter out X The low-energy part of the ray prevents the total dose effect from affecting the test results. 8. The test method for single event soft error with X-rays as claimed in claim 5, characterized in that, after the X-ray radiation ends, the readback data specifically comprises: after the X-ray radiation ends, adjusting the bias voltage to normal The voltage state of the data, and read back all the data, and compare it with the initially filled data. If an error occurs, record the wrong data and the wrong address; then fill the data again, and read back the data to ensure that the During the process, no error occurred, and the SRAM test device can perform normal data reading and writing. The purpose is to ensure that the flip error of the device during irradiation is a single event soft error caused by high-energy electrons, rather than the cumulative total dose. error caused by the effect.