Photoelectric Characterization of Intended Source UTBB 1 T FDSOI Based Image Sensor

This paper presents the photoelectrical performance evaluation of proposed intended sourceUltra Thin Body Bias (UTBB) 1 T-FDSOI based image sensor and also a comparative analysis has done with conventional 1 T-UTBB based image sensor [1]. The photoelectrical conversion in proposed image sensor is achieved by utilizing back biasing in intended source UTBBFDSOI based MOSFET. Inthe proposedintended source Ultra Thin Body Bias (UTBB) 1 T-FDSOI based image sensor,the photoelectrical conversion is done within the substrate region (well region) and that signal is read out by MOSFET placed on SOI substrate. The comparative analysis between proposed image sensor and conventional 1 T-UTBB based image sensor has performed on the basis of various parameters such as back-bias voltage, thickness of buried oxide (BOX) layer and doping concentration. In this investigation the proposed image sensor has found better light sensitivity in contrast with conventional image sensor. The performance evaluation of proposed image sensor has done by TCAD simulations.


Introduction
The advent of silicon-on-insulator (SOI) technology become an important revolution in semiconductor industry due to its proper isolation created by buried oxide (BOX) layer. Moreover, some of its important property like better immunity reduced latch-up issues, lower parasitic capacitances, reduced short channel-effect, higher switching speed, least hot carrier effect provides a better opportunity as compared with conventional MOSFET for higher electrical performances [2]. With addition of BOX layer in SOI MOSFET parasitic capacitances and leakage current is drastically reduced and this property of SOI devices makes its useful in high frequency and low power consumption circuitry. Since Si Waveguide have excellent confinement of optical signal and also have a low loss of optical signal, so important features of SOI substrate make it possible to be used in electronics and photonics circuitry [3]. On the basis of better electricalperformance of SOI MOSFET, it is also find application in image sensor. Today image sensor become an essential needs and find application in space and navigation, consumer electronics (smart phone, digital camera, optical mouse, and in medical equipment) and also in biological inspection [4][5][6]. The performance of traditional image sensor are categorize in two types, first is charge coupled device based image sensror (CCD image sensor), second is complementary metal-oxide-semiconductor image sensor (CIS) [7,8]. The thick depletion region is used in charge coupled image sensor for collection of light signal and also utilizes these region for conversion of photon to electron and then these converted electron transferred to the output terminal. CCD sensors are more advantageous than CIS sensor in terms of certain parameter like high fill factor and high quantum efficiency. These advantages comes because of structural change in CCD like in CCD one pixel cell is comprises of single MOS capacitor with semi-transparent gate, however CIS utilizes one single photodiode and three different transistor, one transistor for reseting the image sensor, one transistor for driving the source follower and one additional transistor for addressing. Because of structural complexity of CIS sensor cell CIS sensor is inferior in high fill factor and high quantum efficiency. Moreover, performance of CIS sensor is much better than CCD sensor since it have reduced noise performance, low power consumption, high speed and also its compatibility with peripheral logic circuit makes it low cast and adaptive for large scale production. There are various researcher proposed different traditional CIS sensor with three-dimensional integration technology so that problem of inferior fill factor and quantum efficiency of conventional CIS sensor has been overcome but this integrational technology has certain limitation like design and process complexity [9]. Another way to get rid of inferior fill factor and low quantum efficiency,a new active pixel image sensor using single transistorhas been proposed where without transferring charge carrier to the output node, A MOSFET which is arranged on top silicon layer is used for sensing the photo generatedelectron [10]. Moreover various structures has been also proposed and presented where single transistor is used to sense the photo-generated carrier, and magnifying within same transistor and then read directly [11][12].
Todays, there is rapid progress in decreasing structure size in semiconductor technology. Therefore, miniaturized image sensor are more viable. Thus, in order to get a miniaturized image sensor, the pixel size of image sensor has been continuously scaled so that better performance of the image sensor is achieved [13][14]. With scaled pixel size of image sensor, it is easy to get high resolution of camera sensor and also the requirement of better image quality is achieved [15].
An ultra-thin body bias (UTBB)based SOI MOSFET has been continuously developing since its gate terminal has better ability to control the channel so it has least off-state current [16].In order to achieve high performance in UTBB FDSOI MOSFET back biasing is used to change the threshold voltage of FDSOI MOSFET placed on top silicon layer above the buried oxide region [17,18]. Again due to interface coupling effect, threshold voltage of the MOSFET is also affected by photo-generated electron gathering at BOX/Si substrate interface [19]. Various researcher demonstrate a new type of image sensor which is basically one transistor image sensor by integrating UTBB devices with photodiode,this photodiode built on SOI substrate is used to collect light flux and due to capacitive coupling effect photo-generated carrier causes threshold voltage shifting in both n-type and p-type field effect transistor [20].
In this paper we proposed and demonstrate an intended source UTBB 1 T FDSOI based image sensor and comparative analysishas been done between conventional 1 T-UTBB based image sensor and intended source UTBB 1 T FDSOI based image sensor (proposed) to evaluate the photoelectric characteristics of both devices. Moreover after finding better sensitivity in intended source UTBB 1 T FDSOI based image sensor (proposed) as compared to conventional 1 T-UTBB based image sensor,its performances dependency on various parameter such as by varying dimension parameter and also by varying terminal voltagesis again analyzed and simulated. In this proposed device when negative back voltage is applied on substrate terminal a depletion layer form below buried oxide (BOX) layer and upon illumination light photon is collected inside depletion region and these photon is converted in electron hole pair by electric field setup inside substrate region. These photo generated electrons are collected at BOX/Si substrate interface and because of interface coupling effect threshold voltage modulation will occur which will change the on-state current and threshold voltage of the n-MOSFET placed on the top Si layer. Now when voltage is applied on gate terminal of the n-MOSFET, drain current of n-MOSFET is read and hence, the intensity of the light signal is sensed and evaluation of light intensity will be done. Thus an intended source UTBB 1 T FDSOI based image sensor (proposed)is able to sense light signal, amplify it properly and then read out the amplified signal.TCAD tool Sentaurus [21] is used to simulate and analyze the performance of ourintended source UTBB 1 T FDSOI based image sensor (proposed). 2D ray tracing method is opted for optical simulation. The analytic low field mobility model, parallel electric field dependency model, Lombardi CVT model, Shockley-Read-Hall (SRH) generation recombination model are used for electrical Simulation.

Proposed Device Structure and Characterization
The cross-sectional schematic for simulation and modeling ofconventional 1 T-UTBB based image sensor and proposed intended source UTBB 1 T FDSOI based image sensor is illustrated in Fig. 1(a) and (b). The proposed device structure is constructed asa conventional 1 T-UTBB FDSOI MOSFET which is placed on SOI substrate. The thickness of top silicon layer is 6 nm and the thickness of buried oxide (BOX) layeris 10 nm. The length of the source and drain region are taken as 200 nm out of which 100 nm (Lgap) length is used for the light exposure on the device which is helpful in collecting more photon within shorter time. Further, the length of the channel region is about 30 nm so the total pixel size of the proposed image sensor is about 430 nm and a p + back bias contact is taken at bottom of the p − substrate region for improving the photo electrical performance of the proposed device.The main structure parameter of both conventional and proposed image sensor is depicted in Table I. In the proposeddevice the p-type substrate region is lightly doped so that a thick depletion layer is formed below the BOX region.
Further, in order to secure low loss of light intensity, light photons are passes only through the large gap region and upon illumination through large gap region the photo-generated electron are collected below BOX/Si substrate interface. Due to this more photon collected inside the substrate region that leads to higher image sensor pixel area.
During light exposer on the device, drain terminal voltage (Vd) is set-up to 0.7 V and back bias terminal voltage (Vb) is setup to −0.7 V. On applying the negative biasing at lightly doped p-type substrate a thick depletion layer and an electric field formed below BOX/Si substrate interface and due to this the absorbed photonsare converted into photo-generated carrier and under the influence of electric field the photogenerated electrons are move towards the interface of BOX/ Si substrate. Further, the electrons are gathered under the BOX/Si substrate interface and due to this electrostatic potential under BOX/Si substrate interfacebecomes negative [1]. Moreover under light illumination condition the more negative charges are collected under BOX/Si substrate interface and because ofelectrostatic coupling [19] more positive charges are collected above the BOX/Si interface region. Furthure, these positive charges are nullified by applying more positive gate voltage and due to this the thresholdvoltage is increases which increases the change in on-state (|ΔI on | = |I on -I dark |) current with increasing light intensity. Fig. 2. shows the change in electrostatic potential under BOX/Si substrate interface region of proposed image sensor. As can be inferred from Fig. 2, under dark condition, the electrostatic potential under BOX/Si substrate interface (just below channel region) is positive whereas after illuminating the light intensity of 100mWatt/cm 2 the electrostatic potential is drop drastically under BOX/Si substrate interface from 0.45 V to −0.5 V. Moreover, with increasing light intensity the collection of photo-electron under BOX/Si substrate interface is increases and due to this the electrostatic potential continuously reduces and become more and more negative. Therefore, with large variation in electrostatic potential the threshold voltage of MOSFET built on the top silicon layeris enhanced. The transfer characteristic of proposed intended source UTBB 1 T FDSOI based image sensor is demonstrated in Fig. 3. This characteristic is obtained by taking device under dark and varying light intensity condition. The light intensity are varied from 100 μW/cm 2 to 1 W/cm 2 . The wavelength of the light taken as 550 nm. As can be inferred from characteristic, the threshold voltage of proposed device increases after light illumination and this is continuously increases with increasing light intensity because of by applying light on proposed device, photo generated carrier in substrate region gets modulated (Enhanced) by back-bias voltage. However, after illumination the change in on-state current (|ΔI on | = |I on -I dark |) of the proposed device isincreases with increasing light intensity [1]. The variation in on-state current and threshold voltage with respect to varying light intensity isillustrated in Fig. 4(a and (b).It is clearly observed from both the plots, the proposed device has increasing |ΔIon| and |ΔVth| with increasing light intensity. The reason behind this result is, the electrostatic potential under the BOX/Si substrate interface is drop drastically with increasing light intensity. Moreover, the |ΔIon| value increases from 0.37 mA/μm to 0.49 mA/μm and |ΔVth| increases from 78 mV to 94 mV with light intensity increasing from 100 μWatt/cm2 to 1 W/cm2. However, the Vth value is extracted at constant Id (drain current) of 10 −8 A/μm. Therefore, it can be observed from the both plots, the change in |ΔIon| and |ΔVth| with varying light intensity demonstrate the light detection property of proposed image sensor [10]. Further, the variation in of proposed image sensor [10]. Furthure the variation in threshold voltage (|ΔVth|) with light intensity and back biasing is specified as light induced voltage shift (LIVS) which is defind by following equation [22].
After investigating the light detection property of proposed image sensor, we decided to comprehensively survey the photo-electrical performance of both proposed and conventional image sensor.Therefore, the input characteristic of both the image sensor under dark and light illumination condition of 1 W/cm 2 is demonstrated in Fig. 5(a) and (b). It can be clearly seen from both plots, after.
illumination the threshold voltage is increases and drain current slightly decreases. The reason behind is, during light illumination upon the image sensor photon will be converted  Further, the direction of electric field under the buried oxide layer is pointed downward due to negative back biasing and because of this electric field, photogenerated electron are collected underBOX/Si substrtae interface.Therefore electrostatic potential decreases and threshold voltage of both image sensor increases and on-state current decreases slightly. Although, after evaluating the change in on-state current (|ΔI on | = |I on -I dark |) in both the plot it can be observed that change in on-state current (|ΔI on | = |I on -I dark |) for proposed device is higher than the conventional device which illustrates the better photoelectrical performance of the proposed device. Furthuremore, the change in on-state current (|ΔI on | = |I on -I dark |) and change in threshold voltage ((|ΔV th |) at various light intensity is alsostudied.The variation in on state current (|ΔIon|) and threshold voltage (|ΔVth|) for both image sensors under various light intensity condition illustrated in Fig. 6(a) and (b). For conventional image sensor |ΔVth| increases from 99.5 mV to 129.5 mV and |ΔIon| increases from 0.157 mA/μm to 0.215 mA/μm when light intensity increases from 100μWatt/cm 2 to 1 W/cm 2 whereas for proposed image sensor |ΔVth| increases from 78 mV to 94 mV and |ΔIon|.
Thus, it can be inferred from Fig. 6 (a) and (b), for proposed device the change in on-state (|ΔIon|) current at each value of light intensity is also higher than the conventional image sensor which shows better sensitivity of proposed image sensor in contrast with conventional image sensor.Moreover the variation in threshold voltage for proposed device under different light illumination condition is least as compared to conventional image sensor.

Impact of Various Parameter on the Performance of Proposed Image Sensor
After investigating photo electrical performance of the proposed device, we have discussed other parameter that affect the performance of the proposed image sensor. These parameter are drain to substrate (back bias) voltage, buried (BOX) oxide thickness and well doping concentration (substrate doping). The discussions are given in below section A, B and C.

Impact of Drain to Substrate (Back bias) Voltage on the Performance of Proposed Image Sensor
The impact of drain to substrate (back bias) voltage on the photo-electrical performance of the proposed image sensor is depicted in Fig. 7

Impact of Buried Oxide (BOX) Thickness on the Performance of Proposed Image Sensor
Another parameter that affect the photo-electrical performance of the proposed image sensor is buried oxide (BOX) thickness which is illustrated in Fig. 8. Theevaluation is done at BOX thickness of 10 nm,14 nm and 18 nm and the incident light intensity taken as constant i.e. 1 W/cm2 for each value of BOX thickness. Therefore, the electrostatic potential under BOX/Si substrate interface does not varied with varying buried oxide thickness but it will provide more back bias modulation (shift in threshold voltage and on-state current) with thinner Buried oxide thickness [1]. The electrostatic potential under Box/Si substrate interface has illustrated in Fig. 8. It can be inferred from plot, for proposed device the electrostatic potential under BOX/Si substrate interface (just below channel region) is approximately constant i.e. − 0.7 V at each value of BOX thickness. Furthuremore, Figs. 9 and 10 described the variation in on-state current and variation in threshold voltage with varying BOX thickness. It can be clearly seen from Figs. 9 and 10, the variation in on-state current (|ΔIon|) and threshold voltage (|ΔVth|) is higher for least BOX thickness (10 nm) but the reduction of TBOX provides improved electrostatic potential and hence improve the shift in Ion and Vt but BOX Thicknesses below 10 nm results in quantum effects such as tunneling and the main function of isolation of the BOX is bypassed. Also, the parasitic capacitances increases and theperformance of the transistor deteriorate hence both proposed and conventional devices provides better photo-electrical performance at least BOX thickness of 10 nm.
Further, the Fig. 9(a) and (b) demonstrates, the conventional image sensor have |ΔVth| value is 55.9 mV for20nm BOX thickness and it is improved to 134.4 mV for reduced BOX thickness of 10 nm. Whereas, for proposed image sensor the |ΔVth| value shifted from 69.6 mV to 94 mV with reducing BOX thickness from 18 nm to 10 nm. Thus both plotdepicted the change in threshold voltage is higher for 10 nm thickness of the BOX layer which explain the light sensitivity is better for 10 nm thickness of the BOX layer. Furthermore, this result also reveals the electrostatic potential under BOX/Si substrate interface has provide more back bias modulation (change in threshold voltage and on-state current)with thinner BOX thickness. Moreover, Fig. 10(a) and (b) describes the change in on-state current (|ΔIon|) for both conventional and proposed device which explain sensitivity performance of the both proposed and conventional devices. It can beclearly seen from Fig. 10(a) ,the conventional device have |ΔIon| is 0.078 mA/μm for 20 nm BOXthickness and it is improved to 0.215 mA/μm for reduced BOX thickness of 10 nm whereas Fig. 10(b) demonstrate,for proposed image sensor the |ΔIon| value  shifted from 0.292 mA/μm to 0.494 mA/μm with reducing BOX thickness from 18 nm to 10 nm. Therefore, it can be analyzed that, Both plot shows the change in onstate current (|ΔIon|) is higher for10nm thickness of the BOX layer. Moreover, the comparative analysis reveals the proposed device have higher variation in on-sate current for each value of BOX thickness in contrast with conventional device thus proposed device have better reliability to be find application in higher sensitive image sensor. Further, the variation in threshold voltage is specified by body factor (BF) which is primarily depends on the thicknesses of gate oxide, BOX thickness, and top silicon thickness, hence for this reason, using a UTBB substrate improves the body factor that will affect the back bias modulation and thus performance of the device will improve fdefind by following equation [22].
Here Vb and Vt is back bias (substrate bias voltage) and threshold voltage of proposed image sensor. It can be clearly observed from above equation, higher variation in threshold voltage ( Vth) is achieved at least value of the buried oxide thickness (BOX) which is also analysed by Fig. 9(b).

Impact of Doping Concentration in P − Substrate Region (Well Region) on the Performance of Proposed Image Sensor
The performance of proposed device also have a dependency upon doping concentration in p − substrate region (well region). The p − substrate region (well region) is mainly used by proposed device for light photon collection and also for conversion of these photon into photo-generated carrier. The performance of both image sensors also depends upon the doping concentration of p − substrate region below the BOX region. The comparative plot for variation in on-state current with different doping concentration of p − substrate region is illustrated in Fig. 11 for both image sensors.
It can be clearly observed from Fig. 11, after light illumination both conventional as well as proposed device shows the decreasing variation in on-state currentwith respect to increasing doping well concentration the reason behind is dependency of depletion layer thickness formed below the BOX/Si substrate interface and doping concentration of the p − substrate region (well region) which is inversely proportional to doping concentration of the p − substrate region (well region).Therefore, with a negative back biasing applied on image sensor a depletion layer formed below the BOX/Si substrate interface and after illuminating the image sensor the photo-generated electron gathered into this depletion layer. Thus, when doping concentration of well region increases the depletion layer thickness is decreases and hencereducing the count of photo-generated electron collection. Therefore it can be clearly analyzed that electron collecting efficiency of image sensor degraded and performance of the image sensor is also degraded. Moreover, From Fig 11 (a) and (b) a comparative analysis demonstrates, besides degrading the performance of both device, proposed device still havehigher |ΔIon| for each value of well doping concentration in contrast with conventional device which explains, higher on-state current variation in proposed device provides better sensitivity of light as compared to conventional image device.

Conclusion
Photo-electric characteristic of proposed image sensor is evaluated in this article. Photo-electric property of proposed device is based on biack-bias modulation caused by photogenerated carier in well (substrate) region. The comparative analysis between proposed and convetional device is also evaluated. The impact of drain to substrate (back bias) voltage, buried oxide (BOX) thickness and well doping concentration (substrate doping) on the performance of both the proposed and conventional device is studied well. Photoelectric property of proposed image sensor based on light induced voltage shift is also shown in this article which is described by variation in threshold voltage shift (|ΔVth|) and variation in on-state current (|ΔIon|) which is continuously increases with increasing light intensity and back biasing. Comparative analysis presents higher sensitivity of proposed device in contrast with conventional device which is analysed by change in on-state current (|ΔIon|) with increasing light intensity thus the comparative result shows in proposed device, variation in on-state current (|ΔIon|) is higher with increasing light intensity in contrast with conventional device. Comparative performance analysis based on varying BOX thickness is also present in this article that reveals the proposed device have higher variation in on-sate current for each value of BOX thickness in contrast with conventional device. Comaparative analysis based on another parameter that is well doping concentration (substrate doping) reveals that proposed device still have higher |ΔIon| for each value of well doping concentration (substrate doping) in contrast with conventional device. Thus proposed device have better reliability to be find application in higher sensitive image sensor. Funding We have not received support from any organization for this submitted work also we have not received any funding for conducting this study.
Data Availability The datasets generated andanalysed during the current study are not publicly available due to privacy but we would like to available thesedata on reasonable request.

Declarations
Ethics Approval All procedures performed in our studies is in accordance with the ethical standards of the institutional and national research committee.

Consent to Participate
We have individually consent to participate for this article.

Consent for Publication
We have individual consent to publish this article, but we have objection to having their data published in a journal article.

Competing Interests
We have certified that we have no affiliations or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.