Effect of Transfer Gate Doping on Full Well Capacity and Dark Current in CMOS Active Pixels

Qian WANG; Jiangtao XU; Zhiyuan GAO; Quanmin CHEN

doi:10.3788/gzxb20225111.1104002

Abstract

In recent years, Pinned Photodiode (PPD) CMOS Image Sensors (CISs) are widely used in consumer electronics, medical, and other fields due to their advantages of high integration, low power consumption and low cost. CMOS active pixels play an important role in CISs. The design of the Transfer Gate (TG) affects image quality, which is related to Full Well Capacity (FWC) and dark current. TG affects the feedforward effect by channel potential. The feedforward effect directly influences FWC as the charges in PPD can flow into Floating Diffusion (FD) by thermal emission. In addition, due to the existence of interface states, dark current generates at the interface of the TG channel, which flows into PPD during the integration period. Several papers have analyzed the influence of TG on FWC and dark current, and have proposed different improvement techniques and designs. When a negative bias is added to TG, the channel is in a state of accumulation, isolating the interface state of the channel from the depletion region of PPD so that dark current is greatly reduced. Furthermore, adopting a negative bias to TG increases the channel barrier, inhibiting the feedforward effect and increasing in FWC. A positive voltage adopted to TG is also beneficial to reduce dark current, but will make FWC decrease. Adjusting the doping length of p-type impurities can change the position of the potential barrier, so that dark charges flow to FD. In this paper, the influence of two types of doped transfer gates, named N+TG and P+TG, on full well capacity and dark current are investigated. Channel potential is affected by the work function difference between TG and substrate. As the barrier height between pinned-photodiode and floating diffusion increases, the feedforward effect is inhibited and the full well capacity increases. On the other hand, the channel in charge accumulation can reduce dark current. To analyze the influence of TG doping on FWC and dark current, a typical 4T-PPD pixel structure is used in this paper. The device consists of a PPD, a “special” TG transistor whose drain is a FD node, and three conventional transistors named Reset Transistor (RST), Source Follower (SF), and Row Select (RS) transistor. The two kinds of TG have the same structure except for different doping types. P-type doping is shared with p+doping used in PMOS transistors, so no additional steps need to be introduced. Device level simulation using Technology Computer Aided Design (TCAD) is performed based on 4T pixels working process, trap model is added to the simulation. The concentration of traps is set to 1 × 10¹⁰ traps·cm^-2 and the capture cross-section to 1 × 10^-14 cm². PPD of two doping types of TG integrates for 10 ms in dark conditions. In addition, the light intensity is set to 2 × 10^-3 W/cm² when testing the FWC of PPD. This paper compares P+ TG and N+ TG under the same channel and substrate doping conditions. FWC and dark current characteristics are simulated when the turn-off voltage (V_{TG_off}) is 0 V. Simulation results demonstrate that the full well capacity of photodiode based on P+TG is 26.5% higher than that of N+TG. The dark current is 0.377 times that of N+TG without negative voltage during the exposure. In practical engineering, a negative voltage is usually applied to N+TG during exposure to obtain good full well capacity and dark current characteristics. The opening characteristics of TG affect image lag, which plays an important role in imaging quality and is usually determined by Charge Transfer Efficiency (CTE). CTE of N+TG is greater than 99.999% at 2.3 V, while P+TG requires 3.0 V. P+TG requires a higher voltage to ensure excellent charge transfer. When the FWC of PPD is high, CTE will be negatively affected, resulting in image lag. At this point, the positive charge pump needs to be introduced to ensure transfer characteristics. Under the simulation conditions in this paper, two doping types of TG have good transfer characteristics at 3.3 V.

Keywords

Charge transfer efficiency CMOS active pixels Dark current Full well capacity Image sensors Photodiodes Simulation

0 Introduction

In recent years，Pinned Photodiode（PPD）CMOS Image Sensors（CISs）are widely used in consumer electronics and other fields due to their high performance and low cost ^［1］. CMOS active pixels play an important role in CISs. The design of the Transfer Gate（TG）affects image quality，which is related to Full Well Capacity（FWC）and dark current ^［2-4］. TG affects the feedforward effect by channel potential. The feedforward effect directly influences FWC as the charges in PPD can flow into Floating Diffusion（FD）by thermal emission ^［5］. In addition，due to the existence of interface states，the dark current generates at the interface of the TG channel，which flows into PPD during the integration period ^［6］.

Several papers have analyzed the influence of TG on FWC and dark current，and have proposed different improvement techniques and designs. A negative bias operation of TG is an effective method to reduce dark current under TG ^［7-9］. When a negative bias is added to TG，the channel is in a state of accumulation，isolating the interface state of the channel from the depletion region of PPD so that dark current is greatly reduced. Furthermore，adopting a negative bias to TG increases the channel barrier，inhibiting the feedforward effect and increasing FWC. A positive voltage adopted to TG is also beneficial to reduce dark current due to the sharing mechanism ^［6］，but will make FWC decrease. To reduce dark current，Ref.［10］has proposed a method to change the position of the potential barrier under TG by adjusting the doping length of p-type impurities so that dark charges could flow to FD. This paper investigates the influence of two types of doped transfer gates，named N+TG and P+TG，on FWC and dark current. P-type doping is shared with p+ doping used in PMOS transistors，so no additional steps need to be introduced. The channel potential is affected by the work function difference between TG and the substrate. A higher potential barrier between the TG channel and PPD can effectively suppress the feedforward effect and increase FWC. In addition，accumulated channels can reduce dark current.

1 Impact of TG doping on FWC and dark current

1.1　Work function influence on TG channel potential

To analyze the influence of TG doping on FWC and dark current，a typical 4T-PPD pixel structure is used in this paper. The device cross section is shown in Fig. 1，which consists of a PPD，a “special” TG transistor whose drain is a Floating Diffusion（FD）node，and three conventional transistors named Reset Transistor（RST），Source Follower（SF），and Row Select（RS）transistor. P+ TG and N+ TG are adopted in this paper. The p-type heavy doping range is in the middle region of TG，0.1 μm away from the TG edge. The n-type heavy doping edge of FD is next to the p+ doped boundary on the right side of TG. The upper left corner of Fig. 1 shows the N+ TG structure，and the two kinds of TG have the same structure except for different doping types. To ensure the smooth transfer of charges from PPD to FD，PPD implantation usually extends slightly under TG to assist charge transfer. At the same time，two p-type implantations are carried out under TG，named TG_PD and TG_FD，for threshold adjustment and anti-punch-through respectively. Before illumination，RST opens to reset PPD and FD. The photo-generated electrons are collected and stored by PPD during illumination. After an exposure period，the photoelectric charges in PPD are transferred to FD by switching on TG，where the signal in the charge domain is converted to the voltage domain. When RS is switched on，the signal is buffered to the column bus by SF.

Figure 1.4T-PPD pixel structure profile of P+ TG and N+ TG

The impact of TG doping on FWC and dark current is achieved by channel potential，which is affected by the work function of TG and the substrate. P+TG is proposed in the paper. Fermi energy level of P+TG is near the valence band. The work function difference（W_ms-p）between P+TG and the p-type substrate is calculated by

W_{m s - p} = \frac{E_{g}}{2} - K T l n (\frac{N_{a}}{n_{i}})

（1）

where E_g is the silicon forbidden bandgap，KTln（N_a/n_i）represents the difference between the central bandgap energy of the substrate and Fermi energy level，K is the Boltzmann constant，T is the absolute temperature，N_a is the doping concentration in the channel region of the substrate，n_i is the intrinsic carrier concentration in silicon at room temperature（300 K）. According to the physics of semiconductors ^［11］，part of the voltage caused by the work function difference and traps falls on the oxide layer，while the other part falls on the channel interface.

- \frac{W_{m s - p}}{q} + \frac{q \times N_{t r a p} \times σ_{s}}{C_{o} \times W \times L} = V_{s p} - \frac{Q_{s p a c e}}{C_{o}}

（2）

where N_trap is the density of traps，the typical value of N_trap is 1 × 10¹⁰ traps·cm^-2，σ_s is the effective capture cross-section（In the simulation，σ_s is set to 1×1^-14 cm²），C_o is the gate capacitance per unit area，C_o = ε_o × ε_r/d，V_sp is the interface potential of P+TG channel，Q_space is the density of interface space charge. Substituting the parameters in Table 1 into Eqs.（1）and（2），the left side of the formula is negative after calculation. The work function difference has a higher effect than traps，and the channel interface is in an accumulated state. Q_space uses the metal-insulator-semiconductor model to calculate according to the formula in the accumulative state ^［11］.

Parameter	Value	Unit
Absolute temperature，T	300	K
Silicon forbidden bandgap， E_g	1.12	eV
Boltzmann constant，K	1.38×10^-23	J/K
Doping concentration in channel region of substrate，N_a	1.5×10¹⁷	cm^-3
Intrinsic carrier concentration，n_i	1.5×10¹⁰	cm^-3
Thickness of oxide layer，d	6	nm
Width of TG，W	1	μm
Length of TG，L	0.6	μm
Dielectric constant，ε_o	8.85×10^-12	F/m
Relative dielectric constant of the silicon，ε_s	11.9	-
Relative dielectric constant of the oxide layer，ε_r	3.9	-

Table 1. The design parameters in this paper

View all Tables

Q_{s p a c e} = \frac{\sqrt[]{2} ε_{s} ε_{o} K T}{q L_{D}} \times e^{(- \frac{q V_{s p}}{2 K T})}

（3）

where L_D is the Debye length，L_D =（（ε_oε_sKT）/（q²N_a））^1/2，ε_s is the relative dielectric constant of silicon. Combined with Eqs.（1），（2）and（3），we can get that V_sp is -0.03 V. Generally，traditional TG is heavily doped with n-type impurities. The Fermi energy level of N+TG is near the conduction band. The work function difference is W_ms-n=-E_g/2－（KT）ln（N_a/n_i）. In the same analysis as above，the interface potential of the N+TG channel，named V_sn，is calculated by using the depletion approximation method，and V_snis 0.66 V. Therefore，the channel potential of P+TG is lower than that of N+TG. A higher barrier between the TG channel and PPD is conducive to inhibiting the feedforward effect.

1.2　TG doping influence on FWC and dark current

FWC is an important parameter of CIS. The feedforward effect is the main reason for the decrease in FWC. Fig. 2 shows the potential diagram along the emission current path with TG off. The lower potential of the channel，the more difficult it is for the electrons in PPD to enter FD through thermal emission. In Ref.［12］，when PPD reaches FWC，there is an equilibrium of three main currents in PPD：the light current I_ph=ƞΦ_ph，the sub-threshold current I_DS（mainly caused by the feedforward effect），and the forward current I_fw of PPD.

Figure 2.Potential diagram along the emission current path of P+ TG and N+ TG

I_{D S} = I_{0} e^{- \frac{V_{B x}}{ν_{t h}}}

（4）

I_{f w} = I_{s a t} (e^{\frac{- V_{F W x}}{ν_{t h}}} - 1)

（5）

I_{0} e^{- \frac{V_{B x}}{ν_{t h}}} + I_{s a t} (e^{\frac{- V_{F W x}}{ν_{t h}}} - 1) = η ϕ_{p h}

（6）

where I₀ is the current coefficient，depending on the Richardson constant，the cross-sectional area of the transfer path（PPD-TG interface），and temperature. The subscript x in the symbols mentioned below represents different doping types of TG，taking n and p respectively. V_B_x is the height of the potential barrier between the TG channel and PPD at the full well，which is represented as V_B_x = V_FW_x-V_s_x here，V_FW_x is the PPD voltage at the full well，and I_satis the reverse saturation current through PPD. According to the analysis in Section 1.1，we know that V_sn is greater than V_sp. Combined with Eqs.（4），（5）and（6），it can be concluded that V_FWp is smaller than V_FWn. In addition，V_Bn<V_Bp can also be obtained. The higher barrier can effectively inhibit the feedforward effect. Ignoring the generation and recombination process of the PPD region，FWC can be approximately expressed as the maximum voltage swing of PPD multiplied by C_PPD：

Q_{F W x} = - (V_{F W x} - V_{p i n}) \times C_{P P D}

（7）

where V_pin is the maximum potential variation of hole and electron quasi-Fermi level ^［13］，in general，the voltage after PD reset is V_pin，Q_FW_x is the FWC of PPD. Obviously，Q_FWn_FWp，the FWC of P+TG is improved compared with that of N+ TG.

In a 4T-PPD pixel，the main dark current contributor is the traps at the silicon-oxide interface under TG ^［6］. Dark current caused by traps can be expressed as Ref.［14］.

$I_{d a r k} = q n_{i} (\frac{σ_{s} ν_{t h} N_{t r a p}}{2})$ （8）

where ν_th is the thermal velocity. Combined with the previous analysis，the P+TG channel is in a state of accumulation，isolating the depletion region of PPD from the interface states below TG. The dark current generated by traps is compensated，resulting in the reduction of dark current.

2 Simulation results

A 4T-PPD is simulated in Technology Computer Aided Design（TCAD）. The same trap model as Ref.［6］is added to the simulation. We set the concentration of traps to 1 × 10¹⁰ traps·cm^-2 and the capture cross-section to 1 × 10^-14 cm²^{［10，15-16］}. PPD of two doping types of TG integrates for 10 ms in dark conditions. In addition，the light intensity is set to 2 × 10^-3 W/cm² when testing the FWC of PPD. This paper compares P+ TG and N+ TG under the same channel and substrate doping conditions. FWC and dark current characteristics are simulated when the turn-off voltage（V_{TG_off}）is 0 V.

To prevent punch-through，L_{TG_PD} and L_{TG_FD} are set to 0.3 μm by default. We set the left side coordinate of TG to 5.0 μm. Given FWC and charge transfer problem，the length of TG is set to 0.6 μm. Fig. 3 shows the simulation cross-section with two doping types of TG. The white line is the depletion zone boundary. As shown in Fig. 3，the N+ TG channel is depleted. For a more intuitive view of the results，Fig. 4 shows the one-dimensional electrostatic potential distribution intercepted by two doping types of TG along the X-X'（X-X' is 0.01 μm away from the channel interface）and along the Y-Y'（Y-Y' is located in the middle of TG），respectively. It is obvious that V_sn of N+TG is greater than zero. V_sp obtained by the simulation is -0.032 V，which is slightly different from the previous theoretical calculation of -0.03 V，due to the simulation deviation. P+TG has a higher potential barrier which can better inhibit the feedforward effect. The potential distribution fully indicates that the channel of P+TG is in the state of hole accumulation. FWC with two types of doped TG is shown in Fig. 5（a）. The values are normalized using FWC of N+TG as the standard. Due to the lower channel potential of P+TG，the ability to suppress the feedforward effect is stronger，and FWC is higher. Compared with N+TG，the FWC of P+TG is increased by 26.5%. Furthermore，Fig. 5（b）shows the number of dark charges integrated by PPD of two types of doped TG in 10 ms，which is normalized based on the dark charges（N_dark）of N+TG. Dark current of P+TG is 0.377 times that of N+TG. According to the above analysis，P+TG has better FWC and dark current characteristics when V_{TG_off} is 0 V. In practical engineering，a negative voltage is usually applied to N+TG during exposure to obtain good full well capacity and dark current characteristics. As shown in Fig. 6，FWC and dark current for two types of TG at various V_{TG_off} are obtained. With the decrease of negative voltage，channel potential decreases，so the feedforward effect is inhibited，and the FWC of two types of TG increases. When V_{TG_off} is less than -0.8 V，the FWC of N+TG is higher than that of P+TG at 0 V. As the negative voltage added to N+TG is less than -0.2 V，the depletion zone of PPD is disconnected from the depletion zone of TG channel，and the channel of N+TG accumulates gradually. Dark current of N+TG reduces with V_{TG_off} decreases. When V_{TG_off} is greater than -0.2 V，dark current of N+TG decreases due to the sharing mechanism. For P+TG，the negative voltage has little effect on the dark current because the channel is accumulated. The introduction of negative voltage can improve the full well capacity and dark current characteristics of N+TG，increasing in power consumption.

Figure 3.Two-dimensional simulation profiles

Figure 4.One-dimensional potential diagrams under TG

Figure 5.Simulation results of FWC and dark charges with two types of doped TG at 0 V

Figure 6.Simulation results of FWC and dark current with two types of doped TG at various V_{TG_off}

The opening characteristics of TG affect image lag，which plays an important role in imaging quality and is usually determined by Charge Transfer Efficiency（CTE）^［17］.

$C T E = \frac{Q_{O U T}}{Q_{F W x}}$ （9）

where Q_OUT is charges transferred from PPD to FD. The TCAD tool is used to study the CTE of P+TG and N+ TG. In the simulation，the charge transfer characteristic with opening voltage（V_{TG_on}）of TG from 0 V to 3.3 V is simulated. As shown in Fig. 7，when V_{TG_on} increases，CTE improves and finally tends to be stable，approaching 100%. Before V_{TG_on} is 3.0 V，the transfer characteristic of N+TG is better than that of P+TG because of the higher channel potential. The charge transfer characteristic is controlled by V_{TG_on}. Specially，the CTE of N+TG is greater than 99.999% at 2.3 V，while P+TG requires 3.0 V. As V_{TG_on}（> 3.0 V）continues to rise，the CTE of N+TG and P+TG shows no difference，suggesting that charges can be completely transferred from PPD to FD. Universally，the opening voltage of TG is 3.3 V，Fig. 8 shows the potential profile with TG on. The channel potential of N+TG is higher than that of P+TG. Moreover，the potential gradually increases from the left side to the right side，and charges can be transferred smoothly along the transfer path. This is the ideal situation for simulation. In actual devices，CTE may be lower than the simulation results because of process deviation and annealing process. When the FWC of PPD is high，CTE will be negatively affected，resulting in image lag. At this point，the positive charge pump needs to be introduced to ensure transfer characteristics. Under the simulation conditions in this paper，two doping types of TG have good transfer characteristics at 3.3 V.

Figure 7.Simulation results of CTE with P+TG and N+TG

Figure 8.One-dimensional potential diagram under TG channels with TG on

3 Conclusion

The influence of TG doping on FWC and dark current is analyzed in this paper. The channel potential of P+TG is lower than that of N+TG because of the work function difference between TG and substrate. The higher barrier inhibits the feedforward effect and increases FWC. On the other hand，the channel of P+TG is in a state of accumulation. The interface state under P+TG is isolated from PPD and dark current decreases. Device level simulation using TCAD is performed，when V_{TG_off} is 0V，the full well capacity of the photodiode based on P+TG is 26.5% higher than that of N+TG，and the dark current is 0.377 times that of N+TG. N+TG can optimize FWC and dark current performance by adding negative voltage. In addition，the CTE of N+TG is greater than 99.999% at 2.3 V，while P+TG requires 3.0 V. The voltage applied to P+TG is higher than that to N+TG for complete charge transfer.

References

[1] E R FOSSUM, D B HONDONGWA. Review of the pinned photodiode for CCD and CMOS image sensors. IEEE Journal of the Electron Devices Society, 2, 33-43(2014).

[2] M SARKAR, B BUTTGEN, A J P THEUWISSEN. Feedforward effect in standard CMOS pinned photodiodes. IEEE Transactions Electron Devices, 60, 1154-1161(2013).

[3] S RIZZOLO, V GOIFFON, M ESTRIBEAU et al. Influence of transfer gate design and bias on the radiation hardness of pinned photodiode CMOS image sensors. IEEE Transactions on Nuclear Science, 61, 3290-3301(2014).

[4] S RIZZOLO, V GOIFFON, M ESTRIBEAU et al. Influence of pixel design on charge transfer performances in CMOS image sensors. IEEE Transactions on Electron Devices, 65, 1048-1055(2018).

[5] SARKAR , MUKUL , BUTTGEN et al. Temperature effects on feedforward voltage in standard CMOS pinned photodiodes. IEEE Transactions on Electron Devices, 63, 1963-1968(2016).

[6] O MARCELOT, V GOIFFON, S RIZZOLO et al. Dark current sharing and cancellation mechanisms in CMOS image sensors analyzed by TCAD simulations. IEEE Transactions on Electron Devices, 64, 4985-4991(2017).

[7] B MHEEN, Y SONG, A J P THEUWISSEN. Negative offset operation of four-transistor CMOS image pixels for increased well capacity and suppressed dark current. IEEE Electron Device Letters, 29, 347-349(2008).

[8] T WATANABE, J H PARK, S AOYAMA et al. Effects of negative-bias operation and optical stress on dark current in CMOS image sensors. IEEE Transactions on Electron Devices, 57, 1512-1518(2010).

[9] W ZHANG, Z WEI, Y WANG et al. Investigation of CMOS image sensor dark current reduction by optimizing interface defect, 1-3(2017).

[10] J XU, Q CHEN, Z GAO. Analysis of transfer gate doping profile influence on dark current and FWC in CMOS image sensors. IEEE Journal of the Electron Devices Society, 9, 27-35(2021).

[11] E LIU, B ZHU, J LUO. The physics of semiconductors(2011).