SCR ESD protection device thermal failure temperature prediction method based on energy threshold value

Abstract

The invention discloses an SCR ESD protection device thermal failure temperature prediction method based on an energy threshold value. The SCR ESD protection device thermal failure temperature prediction method specifically comprises the following steps that 1, thermal failure judgment conditions of an SCR ESD protection device are determined; step 2, calculating an estimated failure temperature Tc of the SCR ESD protection device, obtaining a TLP current pulse amplitude corresponding to the estimated failure temperature Tc, and defining the TLP current pulse amplitude as an estimated failure current If; 3, calculating a failure energy threshold value EF of the SCR ESD protection device; 4, calculating the actual bearing energy E of the SCR ESD protection device; and step 5, failure judgment is carried out based on the failure energy threshold EF of the SCR ESD protection device obtained in the step 3 and the actual bearing energy E of the SCR ESD protection device obtained in the step 4. According to the method, the problem that the thermal failure temperature prediction error is too large in an existing prediction method is solved.

Description

Technical Field

[0001] The present invention belongs to the field of microelectronics and solid state electronics technology, and relates to a method for predicting the thermal failure temperature of an SCR ESD protection device based on an energy threshold. Background Art

[0002] ESD (Electrostatic Discharge) seriously affects the reliability and lifespan of electronic devices. With the shrinking of semiconductor process dimensions and the increase in operating frequencies, the importance of ESD protection technology has become increasingly prominent. As a commonly used ESD protection device, SCR (Silicon Controlled Rectifier) has high current-carrying capacity and fast response characteristics, but its thermal management problem is prominent and it is prone to thermal failure due to high temperatures. Currently, research on SCR ESD protection devices mainly focuses on optimizing their electrical performance indicators such as trigger voltage and holding voltage, and the research on their thermal failure is relatively weak. Traditional thermal failure analysis methods are usually based on two boundary conditions: one is that when the peak temperature of the device reaches the melting temperature of silicon (about 1414 °C), it is considered that the device will fail at this time; in fact, the device often fails due to thermal runaway at a temperature far lower than the material melting temperature. The other is that when the concentration of thermally generated carriers in the device is equal to the background doping concentration, the peak temperature of the device is defined as the thermal failure temperature. This method defines the thermal failure temperature of the device only based on the static equilibrium concentration of carriers, without fully considering the dynamic changes of thermally generated carriers and the actual tolerance of the device, resulting in deviations in the results. In addition, although experimental measurement is intuitive, it has hysteresis, increasing the R & D cost and cycle. Therefore, it is urgent to study the thermal failure mechanism of SCR ESD protection devices, so as to establish a method to predict the thermal failure temperature of the device and improve the device R & D efficiency. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for predicting the thermal failure temperature of an SCR ESD protection device based on an energy threshold, which solves the problem of excessive prediction error of the thermal failure temperature of the device existing in the existing prediction methods.

[0004] The technical solution adopted by the present invention is a method for predicting the thermal failure temperature of an SCR ESD protection device based on an energy threshold, which specifically includes the following steps:

[0005] Step 1, determine the thermal failure determination condition of the SCR ESD protection device;

[0006] Step 2, calculate the estimated failure temperature T of the SCR ESD protection device based on the thermal failure determination condition determined in Step 1 c , obtain the TLP current pulse amplitude corresponding to this estimated failure temperature T through Sentaurus simulation, and define this TLP current pulse amplitude as the estimated failure current I c ; f ;

[0007] Step 3, based on the estimated failure temperature T obtained in Step 2 c Calculate the failure energy threshold E of the SCR ESD protection device F ;

[0008] Step 4, calculate the actual energy E that the SCR ESD protection device can withstand;

[0009] Step 5, based on the failure energy threshold E of the SCR ESD protection device obtained in Step 3 F and the actual energy E that the SCR ESD protection device can withstand obtained in Step 4, perform a failure determination.

[0010] The characteristics of the present invention also lie in:

[0011] In Step 1, the thermal failure determination condition of the SCR ESD protection device is:

[0012] When the concentration of thermally generated carriers in the thermal concentration region of the SCR ESD protection device is equal to the background doping concentration of the SCR ESD protection device, the actual energy borne by the thermal concentration region of the device is greater than or equal to the failure energy threshold E of the device at this temperature F .

[0013] The specific process of Step 2 is:

[0014] Based on the thermal failure determination condition of the SCR ESD protection device determined in Step 1, let the concentration of thermally generated carriers n in the thermal concentration region of the SCR ESD protection device d be equal to the background doping concentration of the SCR ESD protection device. The relationship between the concentration of thermally generated carriers n of the SCR ESD protection device d and the estimated failure temperature T c satisfies Equation (1):

[0015]

[0016] Since n d is known, the estimated failure temperature T of the SCR ESD protection device is obtained through Equation (1) c , and the corresponding TLP current pulse amplitude at this estimated failure temperature T c is obtained in Sentaurus simulation, and this TLP current pulse amplitude is defined as the estimated failure current I f .

[0017] The specific process of Step 3 is:

[0018] When the externally applied TLP pulse width t is in the range of 100 ns to 20 μs, the failure power P of the SCR ESD protection device fThe relationship with t satisfies Equation (2):

[0019]

[0020] where A is the area of the thermal concentration region of the SCR ESD protection device, K is the thermal conductivity of the material of the SCR ESD protection device, ρ is the density of the material of the SCR ESD protection device, C p is the specific heat capacity of the material of the SCR ESD protection device, and T 0 is the ambient temperature;

[0021] Integrating the failure power P of the SCR ESD protection device over time t gives the expression for the failure energy threshold E of the SCR ESD protection device, as shown in Equation (3): f Integrating the failure power P of the SCR ESD protection device over time t gives the expression for the failure energy threshold E of the SCR ESD protection device, as shown in Equation (3): F as shown in Equation (3):

[0022]

[0023] where t is the TLP pulse width.

[0024] The specific process of Step 4 is as follows:

[0025] For the thermal runaway of the SCR ESD protection device, a rectangular black-box heat source model is used for analysis. It is assumed that all the power of the SCR ESD protection device is concentrated in the rectangular black-box heat source, and the peak temperature T of the SCR ESD protection device max is expressed as:

[0026] T max = T 0 + ΔT (4)

[0027] where ΔT is the temperature rise, and the relationship between ΔT and the power P t satisfies the following piecewise function:

[0028]

[0029] where D is the thermal diffusivity of the material of the SCR ESD protection device, ρ is the density of the material of the SCR ESD protection device, t a 、t b 、t c are three time constants, defined as follows:

[0030]

[0031] where a, b, and c are the length, width, and height of the rectangular black-box heat source respectively;

[0032] Equations (5)-(8) reflect the power actually borne by the thermal concentration region of the SCR ESD protection device at different times. Integrating the power P in Equations (5)-(8) over time t gives the energy E actually borne by the SCR ESD protection device, and the expression is as follows: t Integrating the power P in Equations (5)-(8) over time t gives the energy E actually borne by the SCR ESD protection device, and the expression is as follows:

[0033]

[0034] The specific process of Step 5 is as follows: Compare the energy E actually borne by the SCR ESD protection device with the failure energy threshold E f under the predicted failure current I F of the SCR ESD protection device. According to the comparison result, determine whether the SCR ESD protection device fails.

[0035] In Step 5, there are the following three comparison results:

[0036] 1) If, under the predicted failure current I f of the SCR ESD protection device, the energy E actually borne by the SCR ESD protection device is equal to the failure energy threshold E F of the SCR ESD protection device, it is determined that the SCR ESD protection device just fails at this time, and the predicted failure temperature T c of the SCR ESD protection device at this time is the temperature T at which thermal failure actually occurs;

[0037] 2) If, under the predicted failure current I f of the SCR ESD protection device, the energy E actually borne by the SCR ESD protection device is lower than the failure energy threshold E F of the SCR ESD protection device, it is determined that the SCR ESD protection device does not fail at this time;

[0038] 3) If, under the predicted failure current I f of the SCR ESD protection device, the energy E actually borne by the SCR ESD protection device exceeds the failure energy threshold E F of the SCR ESD protection device, it is determined that the SCR ESD protection device fails.

[0039] In Step 5, when the comparison result 2) appears, increase the amplitude of the TLP current pulse. After the amplitude of the current pulse increases, the peak temperature of the SCR ESD protection device will also change accordingly, and the power actually borne at the same moment will also change. Therefore, it is necessary to recalculate the energy E actually borne by the device and compare it with the failure energy threshold E F until the SCR ESD protection device fails;

[0040] When the comparison result 3) appears, reduce the amplitude of the TLP current pulse. After the amplitude of the current pulse is reduced, the peak temperature of the device will also change accordingly, and the actual power carried at the same moment will also change. Therefore, it is necessary to recalculate the energy E actually borne by the device and compare it with the failure energy threshold E F for comparison until the SCR ESD protection device does not fail;

[0041] By processing the above comparison results 2) and 3), determine the TLP current pulse amplitude range where the SCR ESD protection device fails and does not fail; within the TLP current pulse amplitude range where the SCR ESD protection device fails and does not fail, use the bisection method to gradually narrow this range, and repeat the above processing process for comparison results 2) and 3) within the narrowed range until the error between the actually borne energy E and the failure energy threshold E F reaches the expected accuracy, and finally obtain the failure current pulse amplitude I when the SCR ESD protection device undergoes thermal failure; use Sentaurus software to obtain the temperature amplitude T of the SCR ESD protection device corresponding to when the applied TLP current pulse amplitude reaches the failure current pulse amplitude I. At this time, the peak temperature T of the SCR ESD protection device is the actual thermal failure temperature of the SCR ESD protection device.

[0042] The beneficial effect of the present invention is that the present invention comprehensively considers the thermal distribution and carrier behavior of the device, combines the heat generation and heat dissipation mechanisms, and proposes a new method for predicting the thermal failure temperature of the SCR ESD protection device from the perspective of energy threshold. Different from the traditional method that only relies on static temperature or carrier concentration, this method breaks through these limitations and further considers the energy accumulation effect of the device over time under stress, rather than only relying on instantaneous power to measure the bearing capacity of the device. Therefore, this method can more comprehensively and accurately reflect the thermal load of the device and its actual bearing capacity, conforms to the physical mechanism of device failure, and thus provides a more accurate thermal failure prediction. Brief Description of the Drawings

[0043] Figure 1 is a schematic structural diagram of the SCR ESD protection device;

[0044] Figure 2(a) is a lattice temperature distribution diagram of the SCR ESD protection device under a TLP current pulse amplitude of 0.0001 A / μm;

[0045] Figure 2(b) is a lattice temperature distribution diagram of the SCR ESD protection device under a TLP current pulse amplitude of 0.01 A / μm;

[0046] Figure 3(a) is the potential distribution diagram of the SCR ESD protection device under a TLP current pulse amplitude of 0.0001 A / μm;

[0047] Figure 3(b) is the potential distribution diagram of the SCR ESD protection device under a TLP current pulse amplitude of 0.01 A / μm;

[0048] Figure 3(c) is the current density distribution diagram of the SCR ESD protection device under a TLP current pulse amplitude of 0.0001 A / μm;

[0049] Figure 3(d) is the current density distribution diagram of the SCR ESD protection device under a TLP current pulse amplitude of 0.01 A / μm;

[0050] Figure 4 is the flowchart of the method for predicting the thermal failure temperature of the SCR ESD protection device based on the energy threshold of the present invention;

[0051] Figure 5 is the schematic diagram of the hot spot distribution of the SCR ESD device in the method for predicting the thermal failure temperature of the SCR ESD protection device based on the energy threshold of the present invention;

[0052] Figure 6 The change curve of the temperature of the device under different TLP current pulse amplitudes in the method for predicting the thermal failure temperature of the SCR ESD protection device based on the energy threshold of the present invention;

[0053] Figure 7 The power-time curve of the device under the predicted failure current pulse I f of the method for predicting the thermal failure temperature of the SCR ESD protection device based on the energy threshold of the present invention. Detailed implementation manners

[0054] The present invention will be described in detail below with reference to the accompanying drawings and specific implementation manners.

[0055] The method for predicting the thermal failure temperature of the SCR ESD protection device based on the energy threshold of the present invention, as Figure 1 shown, takes the SCR ESD protection device as the research object. The device is provided with a juxtaposed N-type well region and a P-type well region in the P-type substrate, and N+ regions and P+ regions are provided on the surfaces of the N-type well region and the P-type well region, which are respectively connected to the anode and the cathode. The active regions are isolated by shallow trench isolation. When the anode of the device is subjected to ESD stress, the PNP transistor structure composed of the anode P+ region, the N-type well region, and the cathode P+ region (P+ of the N-type well region) serves as the main triggering path to trigger the operation of the device. Subsequently, the PNPN structure composed of the anode P+ region, the N-type well region, the P-type well region, and the cathode N+ region serves as the main path for current discharge.

[0056] To explore the thermal failure mechanism of the device under study, the present invention performed a simulation analysis on the thermal distribution characteristics of the device based on Sentaurus software. First, the thermal distribution characteristics of the device under the action of different TLP (transmission line pulse) current pulse amplitudes were simulated. Figures 2(a) (where the legend in the figure Lattice Temperature refers to the lattice temperature of the device) and 2(b) (where the legend in the figure Lattice Temperature refers to the lattice temperature of the device) respectively show the lattice temperature distribution of the SCR ESD protection device under 0.0001 A / μm and 0.01 A / μm TLP current pulses. As can be seen from the figures, under the action of a smaller TLP current pulse, the high-temperature risk point is located at the J1 junction (the PN junction formed by the cathode P+ region and the N-well region), and under a larger TLP current pulse, the high-temperature risk point shifts to the J2 junction (the PN junction formed by the N-well region and the P-well region). During the triggering process, the J1 junction (the PN junction formed by the cathode P+ region and the N-well region) will experience a relatively high temperature, but this phenomenon only appears near the triggering state and has a relatively small impact on the overall device. In contrast, the temperature distribution of the J2 junction (the PN junction formed by the N-well region and the P-well region) is the key factor determining the thermal characteristics of the device.

[0057] Furthermore, through simulation, the potential distribution and current density distribution of the device under study under 0.0001 A / μm and 0.01 A / μm TLP current pulses were obtained, as shown in Figures 3(a) (where the legend in the figure Electro staticPotential refers to the potential of the device), 3(b) (where the legend in the figure ElectrostaticPotential refers to the potential of the device), 3(c) (where the legend in the figure Abs(TotalCurrentDensity) refers to the current density of the device), and 3(d) (where the legend in the figure Abs(TotalCurrentDensity) refers to the current density of the device) respectively. By comparing and analyzing the potential and current density magnitudes of the J1 junction (the PN junction formed by the cathode P+ region and the N-well region) and the J2 junction (the PN junction formed by the N-well region and the P-well region) under different TLP pulses (transmission line pulses), it can be concluded that the cause of device failure lies in current concentration, and the fundamental cause of current concentration is the local aggregation of carriers.

[0058] Example 1

[0059] The thermal failure temperature prediction method for the SCR ESD protection device based on the energy threshold of the present invention specifically includes the following steps:

[0060] Step 1, determine the thermal failure determination condition of the SCR ESD protection device;

[0061] Step 2, calculate the estimated failure temperature T of the SCR ESD protection device c, the estimated failure temperature T is obtained through Sentaurus simulation c and the corresponding TLP current pulse amplitude at this time is obtained, and the TLP current pulse amplitude is defined as the estimated failure current I f ;

[0062] Step 3, calculate the failure energy threshold E of the SCR ESD protection device F ;

[0063] Step 4, calculate the energy E actually borne by the SCR ESD protection device;

[0064] Step 5, failure determination.

[0065] Embodiment 2

[0066] The specific process of Step 1 is as follows: Based on the above analysis of the thermal failure reasons of the studied SCR ESD protection device, it can be concluded that the root cause of the thermal failure of the SCR ESD protection device is the carrier aggregation inside the SCR ESD protection device. Combining the thermal failure mechanism of the SCR ESD protection device and its actual bearing capacity, the determination condition for the thermal failure of the SCR ESD protection device: When the SCR ESD protection device undergoes thermal failure, after the thermal generation carrier concentration in the thermal concentration region of the SCR ESD protection device is equal to the background doping concentration level of the SCR ESD protection device, the energy E actually borne by the SCR ESD protection device is equal to or exceeds the failure energy threshold E F , thus triggering device thermal failure. Based on the thermal failure determination condition defined in the present invention, the thermal failure temperature of the device is further predicted through theoretical derivation.

[0067] Embodiment 3

[0068] The specific process of Step 2 is as follows: Based on the thermal failure determination condition of the SCR ESD protection device proposed in the present invention (given by Step 1), combined with the specific doping concentration of the SCR ESD protection device, let the thermal generation carrier concentration n in the thermal concentration region of the SCR ESD protection device d be equal to the background doping concentration (specific doping concentration of the research object) level, and calculate the peak temperature T of the SCR ESD protection device at this time c , and define T c as the estimated failure temperature. At this temperature, the relationship between the thermal generation carrier concentration n of the SCR ESD protection device d and the temperature satisfies Equation (1):

[0069]

[0070] According to formula (1), combined with the doping concentration of the selected SCR ESD protection device heat concentration area, the estimated failure temperature T of the SCR ESD protection device can be calculated. c (i.e., the peak temperature inside the device at this time). In the Sentaurus software simulation, there is a one-to-one correspondence between the temperature of the SCR ESD protection device and the amplitude of the external TLP current pulse. Therefore, the corresponding TLP current pulse amplitude at this temperature can be obtained through the Sentaurus simulation software and defined as the estimated failure current I f .

[0071] After obtaining the estimated failure temperature T of the SCR ESD protection device c And the estimated fault current I f After that, it is necessary to calculate the failure energy threshold E of the device under the estimated failure current amplitude pulse. F (Step 3) and the energy E actually borne by the device (Step 4), and the two are compared to determine whether the device has failed.

[0072] Example 4

[0073] The specific process of step 3 is as follows: The thermal breakdown characteristics of the SCR ESD protection device are closely related to the external TLP pulse width. When the external pulse width is in the range of 100ns to 20μs, the failure power P of the SCR ESD protection device is f The relationship between the pulse width t satisfies the expression shown in formula (2):

[0074]

[0075] Where A is the area of ​​the heat concentration region of the SCR ESD protection device (the default device area factor is 1 during simulation), K is the thermal conductivity of the SCR ESD protection device material (silicon), ρ is the density of the SCR ESD protection device material (silicon), C p is the specific heat capacity of the SCRESD protection device material (silicon). c is the failure temperature of the SCR ESD protection device, T 0 is the ambient temperature.

[0076] Power as the failure judgment basis mainly focuses on the magnitude of the instantaneous thermal power. Although it can reflect the power load of the device at a certain moment, it ignores the cumulative effect of power over time in the SCR ESD protection device. In contrast, by using an energy-based failure determination method, the total energy absorbed by the device during the entire SCR ESD protection device pulse is directly quantified through the integration of power over time, thus more comprehensively and accurately reflecting the thermal load and actual tolerance of the SCR ESD protection device, and being more in line with the physical mechanism of SCR ESD protection device failure. Therefore, integrating power over time gives the device failure energy threshold E F The expression of, as shown in Equation (3):

[0077]

[0078] where t is the TLP pulse width. Using the Matlab tool to calculate Equations (2) and (3), the device failure energy threshold E F can be obtained as the standard for subsequent comparison.

[0079] Example 5

[0080] The specific process of Step 4 is as follows:

[0081] For the thermal runaway of the SCR ESD protection device, a rectangular black box heat source model can be used for analysis. This model assumes that all power is concentrated in the rectangular black box heat source, and the peak temperature T max of the device can be expressed as:

[0082] T max = T 0 + ΔT (4)

[0083] where T 0 is the initial temperature of the SCR ESD protection device (i.e., the ambient temperature), and ΔT is the temperature rise, and its relationship with the power P t satisfies the following piecewise function:

[0084]

[0085]

[0086] where K is the thermal conductivity of the SCR ESD protection device material (silicon), C p is the specific heat capacity of the SCR ESD protection device material (silicon), D is the thermal diffusivity of the SCR ESD protection device material (silicon), ρ is the density of the SCR ESD protection device material (silicon), t a , t b , t c are three time constants, respectively defined as follows:

[0087]

[0088] Among them, a, b, and c are the length, width, and height of the rectangular black-box heat source. The rectangular black-box heat source is an idealized model used to describe these hot spot concentration regions and can effectively characterize the thermal effect characteristics of local heat sources. The rectangular black-box heat source refers to a heat source region with a regular shape and a uniform power density distribution, aiming to simulate the actual hot spot region inside the device during an ESD event. Through this model, complex heat distribution problems can be simplified. Among them, the length and height (a, c) of the rectangular black-box heat source can be determined by obtaining the distribution range of the hot spot region of the device through Sentaurus software. The width (b) of the rectangular black-box heat source generally refers to the area factor of the device during simulation (in order to be consistent with the area of the device's heat concentration region in Equation (1), the area factor is usually taken as 1).

[0089] Equations (5)-(8) reflect the power actually borne by the device's heat concentration region at different times. Similarly, integrating the power P in Equations (5)-(8) with respect to time gives the energy E actually borne by the device. The expression is as follows: t Integrating over time gives the energy E actually borne by the device. The expression is as follows:

[0090]

[0091] The temperature rise ΔT involved in Equations (5)-(8) is a parameter that varies with time, and its specific value can be obtained through Sentaurus software. The specific method is as follows: First, use Sentaurus software to obtain the curve of the device's maximum temperature T varying with time under the corresponding TLP current pulse amplitude and export the data to Excel. Then, subtract the ambient temperature T max from it to obtain the temperature rise ΔT at different time points and save the temperature rise data as a file. Next, use the Matlab tool to import this temperature rise file and calculate according to Equations (5)-(13) to obtain the energy E actually borne by the device. 0 from it, so as to obtain the temperature rise ΔT at different time points, and save the temperature rise data as a file. Then, use the Matlab tool to import this temperature rise file and calculate according to Equations (5)-(13) to obtain the energy E actually borne by the device.

[0092] Example 6

[0093] The specific process of Step 5 is as follows: Through Step 3 and Step 5, the device failure energy threshold E f and the energy E actually borne by the device under the predicted failure current I F of the SCR ESD protection device can be obtained, and then the failure determination of the device can be carried out.

[0094] 1) Compare the energy actually borne by the SCR ESD protection device (calculated by Step 4) and the failure energy threshold of the SCR ESD protection device (calculated by Step 3) under the predicted failure current I f .

[0095] 2) If, under the estimated failure current I f , the actual energy E borne by the SCR ESD protection device is equal to the failure energy threshold E F of the SCR ESD protection device, it is determined that the SCR ESD protection device just fails at this time. At this time, the estimated failure temperature T c of the SCR ESD protection device is the temperature T at which the SCR ESD protection device actually undergoes thermal failure.

[0096] 3) If, under the estimated failure current I f , the actual energy E borne by the SCR ESD protection device is lower than the failure energy threshold E F of the device, it is determined that the SCR ESD protection device does not fail at this time. In this case, increase the amplitude of the TLP current pulse. After the amplitude of the current pulse increases, the peak temperature of the SCR ESD protection device will also change, and the actual power borne at the same moment will also change. Therefore, it is necessary to recalculate the actual energy E borne by the device and compare it with the failure energy threshold E F until the SCR ESD protection device fails.

[0097] 4) If, under the estimated failure current I f , the actual energy E borne by the SCR ESD protection device exceeds the failure energy threshold E F of the SCR ESD protection device, it is determined that the SCR ESD protection device fails. At this time, the amplitude of the TLP current pulse can be reduced. After the amplitude of the current pulse decreases, the peak temperature of the device will also change, and the actual power borne at the same moment will also change. Therefore, it is necessary to recalculate the actual energy E borne by the device and compare it with the failure energy threshold E F until the SCR ESD protection device does not fail.

[0098] 5) Through step 3) or step 4), the TLP current pulse amplitude range for the failure and non-failure of the SCR ESD protection device can be determined.

[0099] 6) Within the TLP current pulse amplitude range for the failure and non-failure of the SCR ESD protection device, the bisection method is used to gradually narrow this range, and steps 3) to 5) are repeated within the narrowed range. This process can be carried out repeatedly until the expected accuracy is achieved (the present invention requires that the error between the actual energy borne and the failure energy threshold shall not exceed 0.5%). Finally, the failure current pulse amplitude I when the SCR ESD protection device undergoes thermal failure can be obtained.

[0100] 7) Using Sentaurus software, when the amplitude of the applied TLP current pulse reaches the failure current pulse amplitude I for the SCR ESD protection device, the corresponding temperature amplitude T of the SCR ESD protection device is obtained. At this time, the peak temperature T of the SCR ESD protection device is the temperature at which the SCR ESD protection device actually undergoes thermal failure.

[0101] Example 7

[0102] The characteristic parameters of the SCR ESD protection device (silicon material) required for the prediction method proposed by the present invention are as follows: K = 1.5 (W / cm·K) (thermal conductivity of the semiconductor device material (silicon)), ρ = 2.34 g / cm 3 (density of the semiconductor device material (silicon)), C p = 700 J / (kg·K) (specific heat capacity of the semiconductor device material (silicon)). According to Figure 5 the hot spot distribution map of the selected SCR ESD protection device shown, the area A of the device's heat concentration region can be determined as A = 0.72 μm 2 .

[0103] Based on the calculation using Equation (1), the estimated failure temperature T of the device c = 1000 K. Since there is a one-to-one correspondence between the device temperature and the TLP current pulse amplitude in the Sentaurus simulation software, the temperature change of the device under different TLP current pulse amplitudes can be derived, as Figure 6 shown. Combining the specific value of the estimated failure temperature T c of the SCR ESD protection device, the estimated failure current I C corresponding to the estimated failure temperature T f = 0.05 A / μm can be determined.

[0104] Subsequently, through Equations (2) - (3), the failure energy threshold E of the device F = 4.31×10 -9 J is obtained. Based on the definition of the black box heat source model, the length, width, and height (a, b, c) of the rectangular black box heat source are 0.8 μm, 1.0 μm, and 0.9 μm respectively (the area factor used in this example simulation is 1, so the width of the rectangular black box heat source is 1.0 μm, and the length and height are determined by Figure 5 the heat concentration region range shown)

[0105] Furthermore, according to Equations (9) - (12), the relevant variable D = 9.16×10 -5 m 2 / s (thermal diffusivity of the semiconductor device material (silicon)), t a = 8.687×10 -10 s, tb = 5.560×10 -10 s, t c = 4.257×10 -10 s(t a , t b , t c are three time constants respectively). Combining equations (5) - (12), the variation of the actual power borne by the device with time was calculated, and a power - time curve was generated using Matlab software (as Figure 7 shown). Only the power - time curve of the device under the predicted failure current I f is shown here. The method for generating the power - time curve of the device under different current pulse amplitudes is the same.

[0106] Furthermore, using the Matlab tool to integrate the power - time curve, the actual energy E borne by the device under the predicted failure current I f was obtained as E = 4.7505×10 -9 J. By comparing the device failure energy threshold and the actual energy borne, it was found that the device had failed at this time. To further accurately determine the failure current range, the TLP current pulse amplitude was gradually reduced. When the TLP current pulse amplitude was reduced to 0.048 A / μm, combining equations (5) - (13), the actual energy E borne by the device at this time was calculated as E = 4.2586×10 -9 J, and the device failed. Therefore, the TLP current pulse amplitude range for device failure and non - failure was determined to be [0.048, 0.05]. Within this range, the bisection method was used to further narrow the failure current range. The calculation was repeated until the target accuracy was achieved (in this example, the error between the actual energy borne and the failure energy threshold was required not to exceed 0.5%), and then the repeated process could be stopped. Finally, the TLP current amplitude corresponding to the failure current was determined to be 0.04825 A / μm, and the thermal failure temperature T of the device at this time was 941 K.

[0107] Although there is a deviation of about 0.3% between the actual energy per unit area borne by the device and the theoretical energy per unit area failure threshold, this error is acceptable in engineering applications. Compared with the traditional method based on the silicon melting temperature (1414 °C) as the failure boundary, the accuracy of the thermal failure temperature calculated by this method has been improved by 33.45%. At the same time, compared with the traditional method that only considers the static device thermally generated carrier concentration equal to the background doping concentration (T = 1000 K), the accuracy has been improved by 5.9%. This improvement makes this method more accurate in characterizing the thermal failure characteristics of the device, thus enabling a more precise reliability assessment of the device.

[0108] The prediction method based on energy threshold proposed by the present invention is used to more accurately predict the thermal failure temperature of SCR ESD protection devices. Different from the traditional methods that rely on the silicon melting temperature or the static equilibrium concentration as boundary conditions, this method starts from the perspective of energy threshold and breaks through these limitations. It can more accurately predict the thermal failure temperature of the device, thereby improving the evaluation accuracy of the device reliability. Boundary conditions of two traditional thermal failure analysis methods: one is that when the peak temperature of the device reaches the melting temperature of silicon (about 1414 °C), it is considered that the device will fail; in fact, the device often fails due to thermal runaway at a temperature far lower than the melting temperature of the material, and this method underestimates the risk of device failure. The other is to define the peak temperature of the device corresponding to when the thermally generated carrier concentration of the device reaches the background doping concentration level as the thermal failure temperature of the device. This method defines the thermal failure temperature of the device only based on the static equilibrium concentration, without fully considering the dynamic changes of thermally generated carriers and the actual tolerance of the device, resulting in deviations in the results. The present invention is used to more accurately predict the thermal failure temperature of SCR ESD protection devices. Different from the traditional methods that rely on the silicon melting temperature or the static equilibrium concentration as boundary conditions, this method starts from the perspective of energy threshold and breaks through these limitations. It can more accurately predict the thermal failure temperature of the device, thereby improving the evaluation accuracy of the device reliability.

Claims

1. A method for predicting thermal failure temperature of SCR ESD protection devices based on energy threshold, characterized in that: The specific steps include: Step 1, determining the thermal failure judgment condition of the SCR ESD protection device; Step 2: Calculate the estimated failure temperature T of the SCR ESD protection device based on the thermal failure judgment condition determined in step 1. c , obtain the estimated failure temperature T through Sentaurus simulation c The corresponding TLP current pulse amplitude is defined as the estimated failure current I f ; Step 3: Based on the estimated failure temperature T obtained in step 2 c Calculate the failure energy threshold E of SCR ESD protection devices F ; Step 4, calculating the actual withstand energy E of the SCR ESD protection device; Step 5: Based on step 3, the failure energy threshold E of the SCR ESD protection device is obtained. F The actual withstand energy E of the SCR ESD protection device obtained in step 4 is used to make a failure judgment.

2. The method for predicting thermal failure temperature of SCR ESD protection device based on energy threshold according to claim 1 is characterized in that: In step 1, the thermal failure judgment condition of the SCR ESD protection device is: after the thermally generated carrier concentration in the heat concentration area of ​​the SCR ESD protection device is equal to the background doping concentration of the SCR ESD protection device, the energy actually borne by the device is greater than or equal to the failure energy threshold of the device.

3. The method for predicting thermal failure temperature of SCR ESD protection device based on energy threshold according to claim 2, characterized in that: The specific process of step 2 is: Based on the thermal failure judgment condition of the SCR ESD protection device determined in step 1, let the thermally generated carrier concentration n in the heat concentration area of ​​the SCR ESD protection device be d =Equal to the background doping concentration of the SCR ESD protection device, the thermally generated carrier concentration n of the SCR ESD protection device d and the estimated failure temperature T c The relationship satisfies formula (1): Because n d It is known that the estimated failure temperature T of the SCR ESD protection device can be obtained by formula (1): c , obtain the estimated failure temperature T in Sentaurus simulation c The corresponding TLP current pulse amplitude is defined as the estimated failure current I f .

4. The method for predicting thermal failure temperature of an SCR ESD protection device based on energy threshold according to claim 3 is characterized in that: The specific process of step 3 is as follows: When the external TLP pulse width t is in the range of 100ns to 20μs, the SCR ESD protection device failure power P f The relationship between and t satisfies formula (2): Where A is the area of ​​the heat concentration region of the SCR ESD protection device, K is the thermal conductivity of the SCR ESD protection device material, ρ is the density of the SCR ESD protection device material, C p is the specific heat capacity of the SCR ESD protection device material, T0 is the ambient temperature; The failure power P of SCR ESD protection device f Integrate over time t to obtain the SCR ESD protection device failure energy threshold E F The expression of is shown in formula (3): Where t is the TLP pulse width.

5. The method for predicting thermal failure temperature of SCR ESD protection device based on energy threshold according to claim 4, characterized in that: The specific process of step 4 is as follows: For the thermal runaway of SCR ESD protection devices, a rectangular black box heat source model is used for analysis. It is assumed that all the power of the SCR ESD protection device is concentrated in the rectangular black box heat source. The peak temperature T max It is expressed as: T max =T0+ΔT (4) Among them, ΔT is the temperature rise, ΔT and power P t The relationship satisfies the following piecewise function: Where D is the thermal diffusivity of the SCR ESD protection device material, ρ is the density of the SCR ESD protection device material, t a ,t b ,t c There are three time constants, defined as follows: Among them, a, b, and c are the length, width, and height of the rectangular black box heat source respectively; Formulas (5)-(8) reflect the actual power that the heat concentration area of ​​the SCR ESD protection device bears at different times. The power P of formulas (5)-(8) is t By integrating over time, we can get the energy E actually borne by the SCR ESD protection device, expressed as follows:

6. The method for predicting thermal failure temperature of SCR ESD protection device based on energy threshold according to claim 5, characterized in that: The specific process of step 5 is as follows: Compared with the estimated failure current I f The actual energy E and failure energy threshold E of the SCR ESD protection device under F The size of the SCR ESD protection device is determined based on the comparison result to determine whether it has failed.

7. The method for predicting thermal failure temperature of SCR ESD protection device based on energy threshold according to claim 6, characterized in that: In step 5, there are three comparison results as follows: 1) If the estimated fault current I f Under this condition, the actual energy E borne by the SCR ESD protection device is equal to the failure energy threshold E of the SCR ESD protection device. F , it is determined that the SCR ESD protection device has failed at this time. The estimated failure temperature of the SCR ESD protection device is T c This is the temperature T at which thermal failure actually occurs; 2) If the estimated fault current I f Under this condition, the actual energy E borne by the SCR ESD protection device is lower than the failure energy threshold E of the SCR ESD protection device. F , it is determined that the SCR ESD protection device has not failed at this time; 3) If the estimated fault current I f Under this condition, the actual energy E borne by the SCR ESD protection device exceeds the failure energy threshold E of the SCR ESD protection device. F , it is determined that the SCR ESD protection device has failed.

8. The method for predicting thermal failure temperature of SCR ESD protection device based on energy threshold according to claim 7, characterized in that: In step 5, when comparison result 2) appears, the TLP current pulse amplitude is increased. After the current pulse amplitude increases, the peak temperature of the SCR ESD protection device will also change accordingly, and the actual power carried at the same time will also change. Therefore, it is necessary to recalculate the actual energy E borne by the device and compare it with the failure energy threshold E. F The comparison is performed until the SCRESD protection device fails; When comparison result 3) appears, reduce the TLP current pulse amplitude. After the current pulse amplitude is reduced, the peak temperature of the device will also change. The actual power carried at the same time will also change. Therefore, it is necessary to recalculate the actual energy E borne by the device and compare it with the failure energy threshold E. F The comparison is performed until the SCR ESD protection device does not fail; By processing the comparison results 2) and 3), the TLP current pulse amplitude ranges of the failed and non-failed SCR ESD protection devices are determined; In the TLP current pulse amplitude range between the failed and non-failed SCR ESD protection devices, the binary method is used to gradually reduce the range, and the above-mentioned processing of comparison results 2) and comparison results 3) is repeated in the reduced range until the actual energy E is equal to the failure energy threshold E. F The error between them reaches the expected accuracy, and finally the failure current pulse amplitude I when the SCR ESD protection device fails thermally is obtained; Sentaurus software is used to obtain the temperature amplitude T of the SCR ESD protection device corresponding to the failure current pulse amplitude I when the external TLP current pulse amplitude of the SCR ESD protection device reaches the failure current pulse amplitude I. At this time, the peak temperature T of the SCR ESD protection device is the temperature at which the SCR ESD protection device actually fails thermally.

Images