CN101943721B - Method for fast measuring imprinting effect of ferroelectric film - Google Patents

Abstract

The invention belongs to the technical field of microelectronics, and relates to a test method for imprinting effects of ferroelectric thin films. The present invention quickly measures the imprinting effect of the ferroelectric thin film by measuring the polarization reversal current of the ferroelectric thin film, which includes: (1) After adding a pulse voltage that produces the imprinting effect, immediately add a pulse voltage of opposite polarity to the imprinting voltage And measure the reversal current of the ferroelectric film; (2) Apply a pulse voltage with a preset polarization direction, wait for a period of relaxation time and then add a positive and negative bipolar pulse voltage that acts as an imprint, and then wait for a period of time to generate Imprinting effect, finally add a positive and negative bipolar voltage exactly the same as the previous positive and negative bipolar pulse voltage to measure the reversal current. The invention can replace the traditional method of obtaining Vc by measuring hysteresis loops, can greatly reduce the time required for testing imprinting effects, and has good application prospects.

Description

A Fast Method for Measuring the Imprinting Effect of Ferroelectric Thin Films

technical field

The invention belongs to the technical field of microelectronics, relates to ferroelectric thin film technology and ferroelectric thin film storage device technology, in particular to a method for quickly measuring the imprinting effect of ferroelectric thin film.

Background technique

Ferroelectric thin film memory device is a kind of non-volatile memory device. It uses two different polarization orientations of ferroelectric domains in the electric field as logic units to store data. It has the advantages of fast read and write speed, low driving voltage, high storage density and Due to the advantages of non-volatility and other advantages, it has become a storage device with great potential. At present, it has been firstly applied in electronic products with low power consumption such as mobile phones, walkmans, game cards and digital cameras. Solving the problem of device reliability is the key to the further development of ferroelectric thin film storage devices. As a key content of device reliability, imprinting effect has been widely concerned. The test method of imprinting effect is very important for the study of imprinting effect and device reliability testing. Very important.

The imprinting effect means that the programming pulse experienced by the ferroelectric film is unipolar, then the ferroelectric state corresponding to this polarity will be strengthened, and the hysteresis loop will be shifted along the corresponding voltage axis direction with the coercive voltage Vc (As shown in Figure 1). The imprinting effect will bring two serious problems to ferroelectric thin film storage devices: one is that after a long time of reading and writing operations, the residual polarization value in a certain polarization direction will be reduced, resulting in poor data retention characteristics; the other is that due to A change in the coercive voltage value will cause a change in the reading and writing operation voltage of the device, making the original reading and writing operation voltage invalid.

In the imprinting effect test, there are two mechanisms of adding pulses to produce the imprinting effect, namely, adding a certain time and a certain size of the imprinting voltage to produce the imprinting effect, and adding a preset bias voltage and then adding a reverse pulse and then waiting for different times to produce the imprinting effect The traditional method in the prior art is to measure the hysteresis loop after completing the required pulse process of imprinting effect, and compare the deviation of the hysteresis loop before and after applying the imprinting effect to obtain the change of coercive voltage Vc. The traditional test method has the following defects, because the time of applying the test voltage is too long will introduce a new imprinting effect.

Contents of the invention

The purpose of the present invention is to avoid the shortcomings of long test time and low accuracy of the traditional test method, and provide a method for quickly measuring the imprinting effect of the ferroelectric thin film.

The method for rapidly measuring the imprinting effect of the ferroelectric thin film provided by the present invention is to measure the reversal current after completing the pulse process required for applying the imprinting effect, and calculate the value of the coercive voltage Vc through the reversal current. As shown in Figure 2, since the measurement of the hysteresis loop uses a triangular wave, the triangular wave is composed of many pulses of different heights, the measurement time is at least one second, and the measurement of the reverse current only needs to add a voltage pulse. The time required is only tens of nanoseconds, which can measure the imprinting effect more quickly, and because the measurement of the hysteresis loop takes a long time, the test process itself will cause the imprinting effect, so this method uses the method of measuring the reverse current more accurate.

The method of the invention calculates the coercive voltage Vc by measuring the polarization reversal current of the ferroelectric thin film, and can quickly measure the imprinting effect of the ferroelectric thin film. There are two forms of pulse voltage application to realize this method:

(1) After adding a pulse voltage that produces an imprinting effect, immediately add a pulse voltage of the opposite polarity to the imprinting voltage and measure the reversal current of the ferroelectric film;

(2) Add a pulse voltage with a preset polarization direction, wait for a period of relaxation time and then add a positive and negative bipolar pulse voltage for imprinting, wait for a period of time to produce the imprinting effect, and finally add a positive The negative bipolar pulse voltage is exactly the same as the positive and negative bipolar voltage to measure the reversal current.

In the present invention, as an optional technical solution, the imprinting voltage ranges from -10V to 10V, and the applied time ranges from 50 nanoseconds to 100 seconds.

In the present invention, as an optional technical solution, the magnitude of the pulse voltage for measuring the reverse current is -10V to 10V, and the applied time is 50 nanoseconds to 10 microseconds.

In the present invention, as an optional technical solution, the magnitude of the preset voltage is -10V to 10V, and the applied time is 50 nanoseconds to 10 microseconds.

In the present invention, as an optional technical solution, the relaxation time is 5 seconds to 105 seconds.

In the present invention, as an optional technical solution, the magnitude of the positive and negative bipolar pulse voltage for imprinting is -10V to 10V, and the pulse time is 50 nanoseconds to 10 microseconds.

In the present invention, as an optional technical solution, the imprinting effect generation time between the two positive and negative bipolar pulse voltages is 50 nanoseconds to 100 seconds.

In the present invention, as an optional technical solution, the magnitude of the positive and negative bipolar pulse voltage for measuring the reversal current is -10V to 10V, and the pulse time is 50 nanoseconds to 10 microseconds.

Adopt the testing method that the present invention provides, can significantly reduce the time of testing engraving effect, can test engraving effect in nanosecond level, and more accurate than traditional method (traditional testing method will introduce new engraving because the time of adding test voltage is too long effect), so it has a good application prospect.

Description of drawings

Fig. 1 is a hysteresis loop diagram of a P-V (polarization-voltage) hysteresis loop shifted due to imprinting effect.

Fig. 2 is a schematic diagram of the triangular wave waveform used for testing the hysteresis loop in the traditional test method and the pulse square wave waveform used for testing the reverse current in this patent.

FIG. 3 is a schematic diagram of test waveforms for measuring the imprinting effect produced by applying an imprinting voltage for a certain time and a certain magnitude.

Figure 4 is a graph showing the relationship between the load voltage generated by the inversion current flowing through the load resistance and the pulse voltage time for inverting the electric domain, 100ns and 10s refer to the time for imprinting the voltage.

Figure 5 is a graph showing the relationship between the coercive voltage V _C and the application time t of the imprinting voltage. -5.2V and -2.0V represent two imprinting voltages, and t ₀ is the cut-off point of the imprinting effect.

FIG. 6 is a relationship diagram of the imprinting voltage V _b corresponding to the time boundary point t ₀ when the imprinting effect takes effect.

Fig. 7 is a relationship diagram of the positive and negative coercive voltage ± V _C corresponding to the waiting time t between two positive and negative bipolar pulse voltages, and the waveform diagram in the figure is a schematic diagram of the pulse signal used in the test.

Detailed ways

In the following, there are two mechanisms for adding pulse signals to generate imprinting effects to illustrate the specific implementation methods. The pulse signals required for testing are all edited by the signal generator.

Embodiment 1. Measuring the imprinting effect produced by adding an imprinting voltage of a certain time and a certain size

Use a signal generator to edit a bias voltage V _b that can produce imprinting effect on the ferroelectric film, the magnitude should satisfy |V _b |>|V _C |, and the time t can range from nanoseconds to seconds. Immediately after the end of V _b , add a reverse voltage (V _SW ) in the opposite direction, the pulse time is about tens of nanoseconds, and the magnitude should satisfy |V _SW |>|V _b |, which reverses the domain polarization and thus A polarization reversal current is generated. Finally, use an oscilloscope to read the load voltage generated by the reverse current (V _L =I _SW ×R _L ). According to the formula of reverse current

${\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; SW</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; SW</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; V</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}}}{{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; L</span>}}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Deduce V _C = V _SW - I _SW × R _L = V _SW - V _L , so as to obtain the value of V _C . The test pulse waveform is shown in Figure 3.

The following will be described in conjunction with specific examples. The ferroelectric film capacitor sample used in the present invention is Pt/IrO ₂ /Pb(Zr _0.4 Ti _0.6 )O ₃ (PZT)/IrO ₂ /Pt/Si, wherein Pt is a platinum electrode, IrO ₂ is an iridium oxide electrode, PZT is a ferroelectric material, and the thickness of the ferroelectric film is 140nm.

Figure 4 is the transient load voltage V _L measured after the voltage is applied for 100 ns and 10 s under the bias voltage of V _b = -2.8 V. It can be seen from the figure that after 8 ns there is a peak value generated by capacitor charging , showing a relatively stable voltage platform, and the corresponding voltage value is V _L required for measuring the imprinting effect mentioned above. It can also be seen from the figure that the load voltage value generated by adding 100ns imprinting bias and the reverse current of 10s imprinting bias is different. caused by changes in the recalcitrant voltage V _C .

Figure 5 shows the curves of the coercive voltage V _C calculated by formula (1) corresponding to different imprinting voltage times. The two curves in the figure correspond to two different engraving voltages (-5.2V and -2.0V). A cut-off point t ₀ can be found from the figure. After the time node t ₀ , the logarithmic coordinates of V _C and t have a linear relationship, and before t ₀ the change of V _C with t is very small. Figure 6 shows the relationship between t ₀ and V _b . It can be seen that the value of t ₀ increases with the increase of |V _b |. The above test conclusions are consistent with the results of using the traditional PV (polarization-voltage) hysteresis loop to test the imprinting effect.

Example 2

Measuring the imprinting effect produced by waiting for a period of time after adding a preset bias voltage and then adding a reverse pulse

Different from the imprinting effect produced by adding an imprinting voltage of a certain size and time, the pulse waveform added for the test is shown in Figure 7. First, a positive voltage pulse V _preset is added as the preset voltage, and its magnitude should satisfy | V _preset | > |V _C |, the pulse time is tens of nanoseconds, and the function of the preset voltage is to make the electric domain of the ferroelectric film turn to one direction first. After adding the preset voltage, in order to keep the injected charge in a stable equilibrium state, it is necessary to wait for a period of relaxation time t _rel , the time length is t _rel =5s+20t, where t is the two parameters added to test the imprinting effect later Time interval between positive and negative bipolar pulses. After staying at t _rel , add a positive and negative bipolar pulse voltage (positive pulse first), the magnitude of which is greater than |V _C |, and the pulse time is about tens of nanoseconds. Its function is to use the negative pulse to make the electric domain polarization reversal. After adding the first positive and negative bipolar pulse voltage, wait for a period of time t. Due to the imprinting effect, the coercive voltage will change after the waiting time t (the magnitude of the change is related to the length of the waiting time). Finally, an identical positive and negative bipolar pulse voltage is added, its function is to reverse the domain polarization to generate a reverse current, and calculate the coercive voltage by measuring the reverse current (V _C = V _SW -I _SW ×R _L ＝V _SW -V _L ), V _L is the voltage generated by the reverse current flowing through the load resistance _RL , which can be read directly with an oscilloscope, because here is a pulse of positive and negative polarity, and the positive pulse generates Once the current is reversed, the negative pulse will generate a reverse current again, so the positive and negative coercive voltage (±V _C ) can be measured.

The present invention actually tests the ferroelectric thin film capacitor sample according to the above steps. As shown in Figure 7, the preset pulse voltage is 6.0V, and the pulse time is 80ns. The size of the two positive and negative bipolar pulse voltages is 4.0V, and the pulse time is 80ns, the waiting time t between two bipolar pulse voltages is from 100ns to 10s, it can be seen from the figure that as t increases, the values of +V _C and -V _C also increase, which means It is consistent with the traditional results of measuring imprinting effect with PV hysteresis loop.

Claims (8)

1. a method of measuring ferroelectric thin film stamp effect fast is characterized in that, calculates coercive voltage Vc through measuring Ferroelectric Thin-film Switching Currents, measures ferroelectric thin film stamp effect fast; It comprises the pulse voltage applying method of following form:

(1) add a pulse voltage that produces the stamp effect after, add the pulse voltage of a pulse voltage opposite polarity that produces the stamp effect therewith at once again and measure Ferroelectric Thin-film Switching Currents;

(2) add a pulse voltage that presets polarised direction; Add a pulse voltage that plays the positive negative bipolar property of imprinting after waiting for one period relaxation time; Wait for a period of time again to produce the stamp effect, add at last one with the identical positive negative bipolar property pulse voltage of positive negative bipolar property pulse voltage before to measure Ferroelectric Thin-film Switching Currents.

2. the method for quick measurement ferroelectric thin film stamp effect according to claim 1 is characterized in that, in described (1), the pulse voltage size of said generation stamp effect is-10V to 10V, to add the time be 50 nanoseconds to 100 second.

3. the method for quick measurement ferroelectric thin film stamp effect according to claim 1 is characterized in that, in described (1), the pulse voltage of said measurement Ferroelectric Thin-film Switching Currents size is-10V to 10V, add the time be 50 nanosecond to 10 microsecond.

4. the method for quick measurement ferroelectric thin film stamp effect according to claim 1 is characterized in that, in described (2), the size of the said pulse voltage that presets polarised direction is-10V to 10V, institute add the time be 50 nanosecond to 10 microsecond.

5. the method for quick measurement ferroelectric thin film stamp effect according to claim 1 is characterized in that in described (2), the said relaxation time is 5 seconds to 105 seconds.

6. the method for quick measurement ferroelectric thin film stamp effect according to claim 1 is characterized in that, in described (2), the size of the positive negative bipolar property pulse voltage of said imprinting is-10V to 10V, the burst length be 50 nanosecond to 10 microsecond.

7. the method for quick measurement ferroelectric thin film stamp effect according to claim 1 is characterized in that, in described (2), the time of the generation stamp effect between said two positive negative bipolar property pulse voltages is 50 nanoseconds to 100 second.

8. the method for quick measurement ferroelectric thin film stamp effect according to claim 1; It is characterized in that; In described (2), the size of the positive negative bipolar property pulse voltage of said measurement Ferroelectric Thin-film Switching Currents is-10V to 10V, the burst length be 50 nanosecond to 10 microsecond.

Images