XIANG Chuanfeng, LI Xiaolong, LU Wu,*, WANG Xin, LIU Mohan, YU Xin, CAI Jiao, ZHANG Ruiqin, HE Chengfa, XUN Mingzhu, LIU Haitao, ZHANG Wei, YU Gang, GUO Qi

(1.Key Laboratory of Functional Materials and Devices for Special Environments, Xinjiang Key Laboratory of Electronic Information Material and Device, Xinjiang Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Urumqi 830011, China; 2.School of Physical Science and Technology, Xinjiang University, Urumqi 830046, China)

Abstract: In this paper, the special test structure gated lateral PNP (GLPNP) bipolar transistors, irradiated in different temperatures, were selected to investigate the response mechanism of temperature and dose to radiation damage. The results show that the both of temperature and dose play primary role in determining the dynamic balance of the interface-trap buildup and annealing. Elevating temperature during irradiation can contribute to the increase of degradation at low dose level, and further decrease in temperature will enhance interface-trap buildup at high dose level.

Key words：temperature effect; total ionizing dose effect; interface-trap charge; GLPNP bipolar transistor

1 Introduction

The enhanced low dose rate sensitivity (ELDRS) was first reported by Enlow in 1991 on bipolar transistors[1], threating operational security of space-borne electronics. It is found that the space is an ultra-low dose rate environment between 1 and 10 mrad(Si)/s by the MPTB test[2-3], and enhanced degradation to occur at the ultra-low dose rates (<10 mrad(Si)/s) present challenges for MIL-STD-883/test method 1019 qualified at 10 mrad(Si)/s. Further, some ELDRS-free devices at 10 mrad(Si)/s showed ELDRS at the ultra-low dose rates[4-5]. Therefore, for bipolar devices applied in space, it must be verified whether the ELDRS at the ultra-low dose rates appears.

Considering the long irradiation time is required to examine a part for ELDRS (<10 mrad(Si)/s), some accelerated technique to accurately evaluate the this ELDRS response should be implemented[6-9]. Among the accelerated test methods, the temperature-switched irradiation (TSI) proposed by Lu et al.[9-16]shows promise as a time-saving and conservative accelerated test method to evaluate ELDRS based on the 10 mrad(Si)/s. However, the TSI do not always provide an upgrade bound for the ultra-low dose rate responses of the devices[17-18]. Previous studies have found that temperature, dose rate and total dose are the key factors affecting the TSI accelerated test method[14-15]. In order to select the reasonable experimental conditions, including temperature, dose rate, and dose, this paper shows an understanding of the mechanisms occurring when the elevated temperature irradiation is employed via the comparative test.

In this paper, the gated lateral PNP (GLPNP) bipolar transistors, a kind of specific devices which apply in separating of interface-trap charges (Nit) and oxide trap charges (Not) at the oxide layer under irradiation conditions[19-22], were used to characterize the interface-trap buildup and annealing under various temperature irradiations. These results show that the balance of the interface-trap buildup and annealing plays primary roles in determining the enhanced degradation. Moreover, the test results of this paper provide a technical reference for choosing the appropriate irradiation temperature to further evaluation the ELDRS of bipolar transistors and circuits in ultra-low dose rate environment.

2 Test sample and condition

Fig.1 Schematic diagram of GLPNP bipolar transistors planar structure

The GLPNP bipolar transistors tested in this paper were fabricated by the State Key Laboratory of Analog Integrated Circuit, China Electronics Technology Group Corporation. Those devices have an independent gate electrode which covers the active base region of the PNP transistor, controlling the potential of the interface independently. The cross section of the GLPNP is illustrated in Fig.1.

The irradiation test was carried out on the60Co-γ radiation source in Xinjiang Technical Institute of Physics and Chemistry, Chinese Academy of Sciences. During the irradiation process, the temperature of the test samples is controlled by a special heating system. The test samples accumulate total dose from 0 to 45 krad(Si) at a fixed dose rate of 0.12 rad(Si)/s under different temperature conditions (140 ℃, 120 ℃, 100 ℃, 80 ℃ and 25 ℃). The elevated temperature irradiation dose steps selected are displayed in Table 1. All pins of samples were grounded in irradiation test and the direct-current electrical parameters of the samples were tested using the Keithley 4200-SCS semiconductor parameter test systems after cooling from high temperature to room temperature.

Table 1 Selected dose step of elevated temperature irradiation for GLPNP bipolar transistors

3 Test result and describe

The Gummel curves of the GLPNP bipolar transistors irradiated to the total dose of 45 krad(Si) at different temperatures are shown in Fig.2. With the increase of the total dose, the base currentIBof the GLPNP transistors also increases significantly. While, the collector currentICis almost constant with the increase of the total dose at different temperatures.

Fig.3 shows the gate sweep (GS) characteristic plots of GLPNP bipolar transistors under60Co-γ ray at a fixed dose rate of 0.12 krad(Si)/s at different temperatures. The peak of the base currentIBfor the GS curve is around -10 V, and it is almost constant with the total dose increases. As the total dose increases, the base currentIBcontinues to increase in GS curve. The GS curve shows base currentIBat a total dose of 45 krad(Si) at 140 ℃ less than 120 ℃, 100 ℃, and 80 ℃. Test results indicate that the annealing effect of trap charges was significant when the irradiation temperature exceeds 140 ℃. This means that the irradiation temperature cannot be selected too high for the accelerated evaluation test.Fig.3 also shows the variations of the base currentIB,as the total dose increases, the base currentIBincreases.

Fig.2 Variation of Gummel plot versus total dose of GLPNP bipolar transistors at different temperatures

Fig.3 Variation of GS plot versus total dose of GLPNP bipolar transistors at different temperatures

Fig.4 Variation of normalized current gain versus total dose of GLPNP bipolar transistors at different temperatures

To characterize the ionizing radiation damage of GLPNP bipolar transistors at different temperatures, the normalized current gain is extracted, as shown in Fig.4. As the total dose accumulates, the normalized current gain decreases, saturates or even anneals at higher temperature in GLPNP bipolar transistors. In the low dose level, the normalized current gain degradation at 140 ℃ is the most serious, while in the medium and high dose range, the normalized current gain degradation at 80 ℃ is the most serious. This indicates that the temperature sensitivity of radiation damage of GLPNP bipolar transistors to varies in different dose ranges.

Previous studies have shown that the main cause for the emergence of ELDRS was the accumulation of interface-trap charges in the interface between Si and SiO2[23-25]. The interface-trap charges, generated at the interface between Si and SiO2, cause the growth of the Shockley-Read-Hall (SRH) recombination rate, which increase the base current, when the transistor accumulates total dose. The plot of interface-trap charge versus total dose at different temperatures is extracted through the GS curve, as shown in Fig.5. It is obvious that the dose response of interface-trap charges densities depend on the temperature. Comparing the result from the temperature irradiation at 140 ℃, a greater interface-trap charges density is observed at low dose level, <5 krad(Si), but a saturation trend does occur for a target of high dose level. Similarly, the interface-trap charges increase with the temperature, and then, the formation of interface-trap charges favors in lower temperature. This suggests that the temperature play primary roles in managing the balance of the interface-trap buildup and annealing, with dependent of the dose.

Fig.5 Interface-trap charge at GLPNP bipolar transistors with total dose of 0.12 rad(Si)/s at different temperatures

4 Analysis and discussion

Previous studies have shown that the combination of proton H+in the oxide layer with Si-H dangling bonds near the interface is the key process for the growth of interface-trap charges[26]. For bipolar devices, the protons H+in the oxide layer mainly come from the direct release of hydrogen-containing defects. The process depends on temperature, and the higher the temperature, the more pro-tons H+are released directly in the low dose levels. At low dose level, elevated the temperature can promoted the concentration of proton H+, while in the high dose level, the hydrogen dimerization reaction can be reduced by decreasing the temperature. At high dose level, the competition mechanism between the release of protons H+and the protons H+dimerization reaction is the key factor affecting interface-trap charges density, which leads to the differences in the enhanced degradation of the GLPNP bipolar transistors base currentIBunder different temperature conditions ultimately.

The proton H+transport and interactions at or near the interface for room temperature and elevated temperature are shown in Fig.6. The reaction rate of proton H+release depends on the reaction rate coefficient and the concentration of reactants (VH2and VH). The loss of proton H+also depends on the reaction rate coefficient and the concentration of reactants. Especially at low temperature, the concentration of reactants (VH and proton H+) is the main inhibited factor of proton H+consumption. As shown in Fig.6a, proton H+directly induced by holes under room temperature irradiation is transported to the Si-SiO2interface. Because of the reaction rate factor, defect density and proton H+concentration, proton H+consumption is less, and proton H+depassivates the Si-H bond and forms the interface-trap charge.

Oxide conditions: a—Room temperature; b—Elevated temperature; c—Moderate temperature H+ is a proton, V+ is an oxygen vacancy, VH2 is a doubly hydrogenated oxygen vacancy, VH is a hydrogenated oxygen vacancy, O is a Si-H bond, X is an interface trap Fig.6 Proton transport and interaction at or near interface for room temperature and elevated temperature

With the increase of temperature at low dose level, the reaction rate coefficient of proton H+release reaction increases, the concentration of proton H+increases, and the growth rate of interface-trap charges increase according to Eqs. (1) and (2). In this case, the loss of proton H+was inhibited in the low dose level due to the limitation of reactant concentration. As the total dose accumulates, the concentration of reactants is no longer a factor limiting the loss of proton H+. The dimerization reaction of hydrogen molecules is obvious, the loss of proton H+is accelerated, and the growth of interface-trap charge tends to be saturated. As shown in Fig.6b, the release and loss of proton H+are in a state of dynamic equilibrium.

Kn=Lc×D×exp(-Eb/kBT)

(1)

WhereKnis reaction rate coefficient,Lcis the capture length of the defect,Dis the diffusivity of the diffusing species,Ebis the reaction barrier,kBis the Boltzmann constant, andTis the temperature.

D=D0×exp(-Ed/kBT)

(2)

WhereD0is a constant, andEdis the diffusion energy.

When the temperature is appropriately lowered, the rate of proton H+release reaction and hydrogen dimerization reaction decrease. According to literature[27], the barrier energy of proton H+release is 0.4 eV, and the barrier energy of hydrogen dimerization reaction is 0.8 eV. It can be seen from Eq. (1) that the reaction rate coefficient of hydrogen dimerization reaction is more sensitive to temperature than that of proton H+release reaction, so the hydrogen dimerization reaction is relatively suppressed by lowering the temperature, and the density of the interface-trap charges is further increased.

To effectively evaluate the ELDRS at low dose rates for GLPNP bipolar transistors, elevated temperature irradiation at low dose level, and appropriately reducing temperature irradiation at high dose level can promote the concentration of proton H+and promote the growth of interface-trap charges.

5 Conclusion

In this work, the GLPNP bipolar transistors were used to characterize the degradation of electrical parameters and the distribution of defects after accumulating total dose under different temperatures. The test results show that elevated temperature can promote protons H+release, and interface trap charges density increases significantly with the higher temperature at low dose level. The competition mechanism between the release of protons H+and the protons H+dimerization reaction determines the density of interface-trap charges at high dose level, which leads to the radiation damage of 80 ℃ irradiation was greater than that 140 ℃, 120 ℃ and 100 ℃ irradiations. The results show that elevated temperature during irradiation test can contribute to the increase of degradation at low dose level, and further reducing the irradiation temperature will promote interface-trap charges buildup at high dose level. Moreover, the results of this paper provide a technical reference for the further evaluation of ELDRS of bipolar transistors and circuits in low dose rate environment.