45.PhotoconductorsforX-RayImageDetectors Photoconduc · 2017-10-04 · Multiplexer Scan timing Scanning control Fig.45.2 Schematic diagram that shows some pixels of the active matrix

Photoconduc1125

PartE|45

45. Photoconductors for X-Ray Image Detectors

M. Zahangir Kabir, Safa Kasap

Modern flat-panel x-ray imaging detectors haveplayed an important role in the transition fromanalog to digital x-ray imaging. They capturean x-ray image electronically and hence enablea clinical transition to digital radiography. Thischapter critically discusses the material, transportand imaging detector properties (e.g., dark cur-rent) of several potential x-ray photoconductorsand compares them with an ideal photoconduc-tor for use in direct-conversion imaging detectors.The present chapter also considers various met-rics of detector performances including sensitivity,detective quantum efficiency, resolution in termsof the modulation transfer function, image lagand ghosting; and examines how these metricsdepend on the photoconductor material, and de-tector structure and design.

45.1 X-Ray Photoconductors . .................... 112745.1.1 Ideal Photoconductor Properties ......... 112745.1.2 Potential Photoconductors ................. 112845.1.3 Summary and the Future ................... 1135

45.2 Dark Current Limitations ................... 113645.2.1 Charge Carrier Depletion ..................... 113745.2.2 Steady-State Thermal Generation........ 113745.2.3 Carrier Injection ................................ 1138

45.3 Metrics of Detector Performance . ....... 113945.3.1 X-Ray Sensitivity ............................... 114045.3.2 Detective Quantum Efficiency ............. 114145.3.3 Modulation Transfer Function (MTF) ..... 114345.3.4 Image Lag and Ghosting .................... 1144

45.4 Summary .......................................... 1145

References ................................................... 1145

The flat-panel x-ray image detector (sensor) concept isillustrated in Fig. 45.1 where x-rays passing throughthe patient’s hand are incident on the sensor. The sen-sor converts the incident image to a digital image thatcan be sent to a computer and viewed on a monitor.Research over the past two decades [45.1] has indi-cated that the most practical flat-panel medical digitalradiographic systems (mammography, chest radiogra-phy and fluoroscopy) are those based on a large-areaintegrated circuit or active matrix array (AMA). Flat-panel imagers incorporating active matrix arrays arecalled active matrix flat-panel imagers or AMFPIs.Active matrix arrays using hydrogenated amorphoussilicon (a-Si:H) thin-film transistors (TFTs) have beenshown to be practical pixel-addressing systems for dis-plays [45.2]. These a-Si:H arrays can be convertedinto x-ray-sensitive imaging devices by adding a thick(0:1�1mm) x-ray-detecting medium, either a phosphoror a photoconductor. The physical form of the x-rayAMFPI is similar to a film/screen cassette and thus itwill easily fit into current medical film/screen-basedx-ray imaging systems. The x-ray image is stored anddisplayed on the computer almost immediately after thex-ray exposure. The stored image can be rapidly trans-

mitted to remote locations for consultation and analysis.The dynamic range of recently developed AMFPI sys-tems is much higher than film/screen-based imagingsystems [45.3]. AMFPIs are able to read out an entireimage in 1=30 s, sufficiently rapid to perform fluo-roscopy (real-time imaging) [45.4].

The active matrix array used for pixel addressingand readout in an AMFPI consists of many single pix-els, each of which represents a corresponding pixel ofthe image. Each pixel collects charge proportional tothe x-ray radiation that it receives. To generate thissignal charge, either a phosphor converts the x-raysto visible light, which in turn is detected with a pinphotodiode at the pixel (indirect) or an x-ray photo-conductor converts the incident x-rays to charge (direct)in one step. For both the indirect and direct conversionapproaches, the latent image is a charge distribution re-siding on the array’s pixels. The charges are read outby scanning the arrays row by row using peripheralelectronics and multiplexing the parallel columns intoa serial digital signal. This signal is then transmittedto a computer system for storage and display. Severalmanufacturers and academic researchers have used theindirect approach [45.5, 6]. However, we believe that

© Springer International Publishing AG 2017S. Kasap, P. Capper (Eds.), Springer Handbook of Electronic and Photonic Materials, DOI 10.1007/978-3-319-48933-9_45

PartE|45

1126 Part E Novel Materials and Selected Applications

X-rays

X-ray source Object

Flat panelX-ray imagedetector

Computer

Fig. 45.1 Schematic illustration of an active matrix flat-panel imaging (AMFPI) system used for x-ray imaging

the direct approach, due to its higher resolution, has thepotential to produce systems superior in image qual-ity to indirect conversion sensors and be both easierand cheaper to manufacture due to their simpler struc-ture. This chapter considers only the direct-conversionx-ray imager and how its dark current, sensitivity, reso-lution and detective quantum efficiency (DQE) dependon the photoconductor material and detector structure.This chapter also discusses essential photoconductorproperties, charge transport and imaging properties ofpromising photoconductors.

Active matrix arrays allow a monolithic imagingsystem of large area (e.g., 40�40cm) to be constructed.As for conventional integrated circuits, planar process-ing of the array through deposition and doping oflithographically masked individual layers of metals, in-sulators and semiconductors implement the design ofactive matrix arrays. In the future, even larger areasshould become feasible if required. Millions of indi-vidual pixel electrodes in the matrix are connected, asshown in Fig. 45.2. Each pixel has its own thin-filmtransistor (TFT) switch and storage capacitor to storeimage charges. The TFT switches control the imagecharge so that one line of pixels is activated electroni-cally at a time. Normally, all the TFTs are turned off,permitting the latent image charge to accumulate onthe array. The readout is achieved by external elec-tronics and software controlling the state of the TFTswitches. The active matrix array consists of M �N(e.g., 2480� 3072) storage capacitors Cij whose storedimage charge can be read through properly addressingthe TFT (i,j) via the gate (i) and source (j) lines. Thecharges read on each Cij are converted to a digital im-age. The readout is self-scanning in that no external

TFT

Gate line (i+1)Storagecapacitance

Gate line (i)

Gate line (i–1)

Pixel

Electrode (B)

Date (source lines)

S

G

D

i,j+1

(j–1) (j+1)(j)

i, ji, j–1

Computer Digitizer

Multiplexer

Scan

tim

ing

Scan

ning

con

trol

Fig. 45.2 Schematic diagram that shows some pixels ofthe active matrix array for use in x-ray AMFPI withself-scanned electronic readout. The charge distribution re-siding on the panel’s pixels are read out by scanning thearrays row by row using peripheral electronics and multi-plexing the parallel columns into a serial digital signal

means such as a laser are used. The scanning is partof the AMFPI electronics and software, thus permittinga truly compact device.

To construct a direct-conversion x-ray AMFPI,a large-bandgap (> 2 eV) high-atomic-number semi-conductor or x-ray photoconductor (e.g., stabilizedamorphous selenium, a-Se) layer is coated onto theactive matrix array. An electrode (A) is subsequently de-posited onto the photoconductor layer to facilitate theapplication of a biasing potential and, hence, an elec-tric field F in the photoconductor layer as shown inFig. 45.3. The biasing potential applied to the radiation-receiving electrode A (top electrode) may be positiveor negative, the selection of which depends on manyfactors discussed later in this chapter. The applied biasvaries from a few hundred to several thousand V. Thecapacitance Cpc of the photoconductor layer over thepixel is in series with, and much smaller than, the pixelcapacitance Cij attached to the pixel electrode, so thatmost of the applied potential is dropped across the pho-toconductor and not across the pixel capacitance. Theelectron–hole pairs (EHPs) generated in the photocon-ductor by the absorption of x-ray photons travel alongthe field lines and are collected by the electrodes. Ifthe applied bias voltage is positive, then electrons col-lect at the positive bias electrode and holes accumulateon the storage capacitor Cij. Each pixel electrode car-

Photoconductors for X-Ray Image Detectors 45.1 X-Ray Photoconductors 1127Part

E|45.1

–

–+

+

X-raysTop electrode (A)

X-rayphotoconductor

Pixelelectrode (B)

Storagecapacitor Cij

SiO2Gate(Al)

FET channel

D

F

Ground

Glass substrate

Electrostaticshield

S

Fig. 45.3 Cross section of a single pixel .i; j/ with a TFTswitch. The top electrode (A) on the photoconductor isa vacuum-coated metal (e.g., Al). The bottom electrode (B)is the pixel electrode that is one of the plates of the storagecapacitance (Cij). (Not to scale – the TFT height is highlyexaggerated)

ries an amount of charge Qij proportional to the amountof incident x-ray radiation in the photoconductor layerover that pixel, which can be read during self-scanning.

For a historical perspective on the development of a-Se based flat panel detectors, the reader is referred toa number of early papers by Rowlands and cowork-ers [45.7–13] and Lee et al. [45.14, 15].

The selection of the photoconductor material foruse in direct-conversion x-ray image detectors is cur-rently an important research field in electronic materi-als. There are various competing semiconductors, suchas amorphous (a-)Se, HgI2, CdZnTe, PbI2, PbO andorganic perovskites. Detectors based on a-Se have al-ready been developed and successfully commercializedfor mammography [45.16]; the majority of modern di-rect conversion mammographic detectors now use ana-Se alloy as a photoconductor. This chapter discussescharge-transport and imaging detector properties (e.g.,dark current and image lag) of these photoconductorsand compares them with an ideal photoconductor forx-ray imaging detectors. The present chapter also exam-ines various imaging characteristics of photoconductor-based x-ray AMFPIs, including dark current, sensitivity(S), detective quantum efficiency (DQE), resolution interms of the modulation transfer function (MTF), imagelag and ghosting. We examine how these characteris-tics depend not only on the photoconductor’s charge-transport properties but also on the detector structure,i. e., the size of the pixel and the thickness of the photo-conductor.

45.1 X-Ray Photoconductors

45.1.1 Ideal Photoconductor Properties

The performance of direct-conversion x-ray detectorsdepends critically on the selection and design of thephotoconductor. It is therefore instructive to identifywhat constitutes a nearly ideal x-ray photoconductor toguide a search for improved performance or a better ma-terial. Ideally, the photoconductive layer should possessthe following material properties:

1. Most of the incident x-ray radiation should be ab-sorbed within a practical photoconductor thicknessto avoid unnecessary patient exposure. This meansthat, over the energy range of interest, the absorp-tion depth ı of the x-rays must be substantially lessthan the photoconductor layer thickness L. In otherwords, the quantum efficiency (�) should be high.

2. The photoconductor should have high intrinsicx-ray sensitivity, i. e., it must be able to gener-ate as many collectable (free) electron–hole pairs(EHPs) as possible per unit of incident radiation.This means the amount of radiation energy required,denoted by W

˙, to create a single free electron–

hole pair must be as low as possible. Typically,W

˙increases with the bandgap Eg of the photo-

conductor [45.17] and thus a low Eg is desired formaximum x-ray sensitivity.

3. There should be little bulk recombination of elec-trons and holes as they drift to the collectionelectrodes; EHPs are generated in the bulk of thephotoconductor. Bulk recombination is proportionalto the product of the concentration of holes andelectrons, and typically it is negligible for clini-cal exposure rates (i. e., provided the instantaneousx-ray exposure is not too high) [45.18].

4. There should be negligible deep trapping of elec-trons and holes, which means that, for both elec-trons and holes, the schubweg ��F must be muchgreater than L, where � is the drift mobility, � isthe deep-trapping time (lifetime), F is the electricfield and L is the photoconductor layer thickness.The schubweg is the mean distance a carrier driftsbefore it is trapped and unavailable for conduction.The temporal responses of the x-ray image detec-tor, such as lag and ghosting, depend on the rate ofcarrier trapping.

PartE|45.1


5. The diffusion of carriers should be negligible com-pared with their drift. This property ensures lesstime for lateral carrier diffusion and leads to betterspatial resolution.

6. The dark current should be as small as possible,because it is a source of noise. The applied fieldcauses a current to flow through the detector in theabsence of irradiation, which is called the dark cur-rent. This means the contacts to the photoconductorshould be noninjecting and the rate of thermal gen-eration of carriers from various defects or states inthe bandgap should be negligibly small (i. e., darkconductivity is practically zero). Small dark con-ductivity generally requires a wide-bandgap semi-conductor, which conflicts with condition (2) above.The dark current should preferably not exceed10�1000pA=cm2, depending on the clinical appli-cation [45.19].

7. The longest carrier-transit time, which depends onthe smallest drift mobility, must be shorter than theimage readout time and interframe time in fluo-roscopy.

8. The properties of the photoconductor should notchange with time because of repeated exposure tox-rays, i. e., x-ray fatigue and x-ray damage shouldbe negligible.

9. The photoconductor should be easily coated ontothe AMA panel (typically 30� 30 cm and larger),for example by conventional vacuum techniqueswithout raising the temperature of the AMA to dam-aging levels (e.g., � 300 ıC for a-Si panels). Thiseliminates the possibility of using single-crystalmaterials that would require extended exposure tomuch higher temperature if they were to be growndirectly onto the panel.

10. The photoconductor should have uniform charac-teristics over its entire area.

11. The temporal artifacts such as image lag and ghost-ing should be as small as possible. Lag is thecarryover of image charge generated by previousx-ray exposures into subsequent image frames. Theresidual signal fractions following a pulsed x-rayirradiation are referred to as image lag. Ghostingis the change of x-ray sensitivity of the x-ray im-age detector as a result of previous exposure toradiation. In the presence of ghosting, a shadow im-pression of a previously acquired image is visible insubsequent uniform exposures.

The large-area-coating requirement in (9) rules outthe use of crystalline semiconductors, whose only prac-tical production process is to grow large boules, whichare subsequently sliced. Thus, only amorphous or poly-crystalline (poly)photoconductors are currently prac-

tical for use in large-area x-ray imaging detectors.Amorphous selenium (a-Se) is one of the most highlydeveloped photoconductors for large-area detectors dueto its commercial use in photocopiers and laser print-ers as an electrophotographic photoreceptor [45.20]. Infact, the direct-conversion flat-panel imaging technol-ogy has been made possible by the use of two keyelemental amorphous semiconductors: a-Si:H (used forTFTs) and a-Se (used for photoconductor layers). Al-though their properties are different, both can be readilyprepared in large areas, which is essential for an x-rayimaging detector. Stabilized a-Se (a-Se alloyed with0:2�0:5% As and doped with 5�20 ppm Cl) is cur-rently the preferred photoconductor for clinical x-rayimage sensors, because it can be quickly and easilydeposited as a uniform film over large areas and ithas an acceptable x-ray absorption coefficient, goodcharge-transport properties for both holes and electrons,and lower dark current than many competing poly-crystalline layers [45.17, 19]. Flat-panel x-ray imagedetectors with an a-Se photoconductor have been shownto provide excellent images.

There has been active research to find potentialx-ray photoconductors to replace a-Se in flat-panel im-age detectors because of the substantially higher W

˙

and operating electric field of a-Se compared to otherpotential x-ray photoconductors [45.19, 21]. For ex-ample, the typical value of the electric field used ina-Se devices is 10V=�m where the value of W

˙is

about 45 eV; the value of W˙

is 5�6 eV for polycrys-talline mercuric iodide (poly-HgI2) and poly-CdZnTe.The main drawback of polycrystalline materials is theadverse effects of grain boundaries in limiting chargetransport and the nonuniform response of the sen-sor due to large grain sizes. Grain boundaries in thepolycrystalline material are expected to create trappinglevels within the bandgap and introduce potential bar-riers between neighboring grains [45.22, 23]. Anotherdisadvantage of these polycrystalline detectors is thehigher dark current compared to a-Se detectors. How-ever, there have been efforts to improve the materialproperties and reduce the dark currents of poly-HgI2and poly-CdZnTe-based image detectors [45.24, 25].Indeed, experiments on large-area HgI2, PbI2, CdZnTe(< 10% Zn), and PbO polycrystalline x-ray photocon-ductive layers deposited on active matrix arrays haveshown encouraging results [45.26–28]. A more detaileddescription of these three potential photoconductors fordirect-conversion AMFPIs is presented below.

45.1.2 Potential Photoconductors

The properties of an ideal photoconductor for x-ray im-age detectors have been discussed in the section above.


E|45.1

In this section, important properties of potential photo-conductors for x-ray image detectors are discussed andcompared with the ideal case.

Amorphous Selenium (a-Se)Stabilized a-Se can be easily coated as thick films (e.g.,100�1000�m) onto suitable substrates by conventionalvacuum deposition techniques and without the needto raise the substrate temperature beyond 60–70 ıC(much below the damage threshold of the AMA, e.g.,� 300 ıC for a-Si:H panels). Its amorphous state main-tains uniform characteristics to a very fine scale overlarge areas. Stabilized a-Se, not pure a-Se, is used inx-ray sensors, because pure a-Se is thermally unstableand crystallizes over several weeks or months followingmanufacture. Crystalline Se is unsuitable as an x-rayphotoconductor because it has a much lower dark resis-tivity and hence orders of magnitude larger dark currentthan a-Se. Alloying pure a-Se with As (0:2�0:5% As)greatly improves the stability of the composite film andhelps to prevent crystallization. However, it has beenfound that the addition of As has an adverse effect onthe hole lifetime because the arsenic introduces deephole traps. If the alloy is doped with 5�20 parts permillion (ppm) of a halogen (such as Cl), the hole life-time is restored to its initial value. Thus, an a-Se filmthat has been alloyed with 0:2�0:5%As (nominal 0:3%As) and doped with ppm-levels of Cl is called stabilizeda-Se. The density of a-Se is 4:3 g=cm3 with a bandgap(usually called the mobility gap in amorphous semicon-ductors) of Eg � 2:1�2:2 eV.

There are localized states within the mobility gapof a-Se. Some of these are located near the band edges(shallow traps), while some are located deep in theenergy band (deep traps). Localized states are simplytraps and are not extended throughout the material, butare localized in space (both shallow and deep) due tovarious structural defects that are stable at room tem-perature. Drift of both electrons and holes involvesinteractions with shallow and deep traps, as shown inFig. 45.4. Shallow traps reduce the drift mobility; deeptraps prevent the carriers from crossing the photocon-ductor. The effective drift mobility � of carriers is themobility �0 in the extended states reduced by the trap-ping and release events due to the presence of shallowtraps,

� D �c

�c C �r�0 ; (45.1)

where �c and �r are the average capture and releasetimes in the shallow-trap centers. The capture time rep-resents the mean time that a mobile carrier drifts in theextended states before becoming trapped in a shallow-trap center. The release time is the mean time that

Shallowelectrontraps

Shallowholetraps

Deepholetraps

Conductionband

Valenceband

Deepelectron

traps

Fig. 45.4 Diagram illustrating the bandgap of a photocon-ductor with an applied electric field that tilts the bands,encouraging drift of holes in the direction of the field andelectrons counter to the field. Drift of both electrons andholes involves interactions with shallow and deep traps.Shallow traps reduce the drift mobility and deep traps pre-vent the carriers from crossing the photoconductor

a carrier remains in a trap before being released backinto the extended states. Re-emission from a shallowtrap is mostly dominated by thermally activated pro-cesses. The shallow-trap release time is relatively shortand a typical carrier may experience many shallow cap-ture and release events while traversing the detectorthickness (100�1000�m).

Although the nature of the shallow traps in a-Se hasnot been fully established, the drift mobilities of bothholes and electrons are quite reproducible. The room-temperature effective hole mobility �h up to moderatefields (F 10V=�m) is independent of the prepara-tion of the sample and has a value of � 0:12 cm2=Vs,whereas the effective electron mobility �e is in therange 0:003�0:006 cm2=Vs [45.21]. The hole drift mo-bility does not change with the addition of As or Cl.The value of �e decreases with the addition of As to a-Se (e.g., in stabilized a-Se) but Cl doping does not affectit.

Once a carrier is caught in a deep trap, it will re-main immobile until a lattice vibration gives it enoughenergy to be excited back into the extended states,where it can drift once again. The deep-trap releasetime is very long (minutes to hours), and a deeplytrapped carrier is essentially permanently removed fromconduction. Therefore, the carrier lifetime depends onthe concentration of deep rather than shallow traps.The charge-carrier lifetimes vary substantially between

PartE|45.1


different samples and depend on factors such as thesource of a-Se material, impurities, and the prepara-tion method. The electron lifetime �e is particularlysensitive to impurities in the a-Se source material. Thehole lifetime �h drops rapidly with decreasing substratetemperature (temperature of the a-Se substrate duringthe evaporation process) whereas �e is unaffected. In-creasing the As concentration in a-Se decreases �h andincreases �e [45.29]. On the other hand, Cl doping in-creases �h and decreases �e. The typical lifetimes instabilized a-Se are in the range 10�500�s for holes and100�1500�s for electrons [45.21].

The fractional increase in the �e with As addi-tion is greater than the drop in �e. Thus the electronrange (�e�e) increases with As content. The effect of Cldoping on the carrier ranges (�� products) is more pro-nounced than that of As doping. Thus, we can controlboth electron and hole ranges by appropriately choosingthe relative amounts of As and Cl in a-Se.

The electron–hole-pair creation energy W˙

ina-Se has a strong dependence on electric field F butonly a weak dependence on the x-ray photon energyE [45.30, 31]. The quantityW

˙is decreased by increas-

ing either F or E. W˙

at a given E in a-Se follows anempirical relation given by

W˙

� W0˙

C B.E/

Fn; (45.2)

where B.E/ is a constant that depends on E, W0˙

isthe saturated EHP creation energy (at infinite F), andn is typically in the range 0.7–1 [45.32]. The value ofW0

˙

should be 2:2Eg CEphonon [45.33], where Ephonon isa phonon energy term. With Eg � 2:1 eV and Ephonon <0:5 eV, we would expect that W0

˙

� 5 eV. The ener-getic primary electron generates many EHPs but onlya certain fraction of these are free to drift and therest recombine before they can contribute to the pho-tocurrent. There are various possible explanations forthe F dependence of the EHP creation energy. First,the simultaneously generated electron and its hole twinare attracted to each other by their mutual Coulom-bic force and may eventually recombine. This type ofrecombination is called geminate recombination (fromGemini – the twins). Another possible mechanism iscolumnar recombination that involves the recombi-nation of nongeminate electrons and holes generatedclose to each other in the columnar track of the singlehigh-energy electron (primary) created by the absorp-tion of an x-ray. In both the geminate and columnarcases, the number of carriers escaping recombinationshould increase with increasing F, which acts to sep-arate the oppositely charged carriers. The question ofwhether the F dependence of W

˙in a-Se is domi-

nated by geminate or columnar recombination has not

been fully resolved and is currently an area of activeresearch [45.33, 34]. However, the energy dependenceof W

˙is better understood. It decreases slowly with

increasing photon energy in the diagnostic [45.30] andmegavolt range [45.35]. The total change in W

˙from

20 keV to 6MeV is of the factor of 3. This appears to bedue to a reduction of recombination with increase in en-ergy. The rate of deposition of energy per unit distancetraveled by a primary electron decreases as a functionof energy, decreasing the density of EHPs in the col-umn around it. This is expected to reduce columnarrecombination – as is seen. Thus, it appears that at lowenergies the contribution from columnar recombinationis approximately twice that from geminate, but at highenough energy the columnar effect is reduced to zero.The typical field value used in a-Se devices is 10V=�mwhere the value W

˙is 35�55 eV over the diagnostic

beam energy (12�120 keV) and � 15 eV at megavoltenergies.

The dark resistivity of a-Se is often quoted roughlyas 1014�cm. However, experimental results in the liter-ature have shown that the main source of the dark cur-rent is the injection of holes from the metal contacts andthe bulk thermal generation current is negligible [45.36,37]. It was the development of multilayer p-i-n typedevice structures that eventually reduced the dark cur-rent to very low levels (< 0:1 nA=cm2 at F as highas 20V=�m [45.38]) and made an a-Se commerciallyviable detector [45.39]. A pin (p-like-intrinsic-n-like) a-Se structure is shown in Fig. 45.5, where the thicknessof both the p-like and n-like blocking contact layers isa few microns. These p- and n-like-layers are appropri-ately doped to serve as unipolar conducting layers thateasily trap electrons and holes respectively, but allowthe transport of oppositely charged carriers. The rateof emission of these deeply trapped carriers is so smallthat there is no significant current injection into the bulka-Se layer. The commercial a-Se detector (in this exam-ple, the Analogic detector) is n-i-p or p-i-n type wherethe p-like layer is usually a-As2Se3 and the n-like layeris alkali metal doped a-Se [45.36]. Recent work has

–

+++++

––– –

Idark = Minimum

n-type-layer

V

a-Se

p-type-layer

F

x

Fig. 45.5 A multilayer pin-type a-Se device structure forblocking the dark current


E|45.1

shown that an n-i structure with a cold deposited n-likelayer also has a very low dark current [45.39]. Withthese n- and p-like layers, the field at the metal elec-trodes is sufficiently small to minimize charge injectionfrom the contacts, which substantially reduces the darkcurrent.

The image lag in a-Se detectors is under 2% after33ms and < 1% after 0:5 s in the fluoroscopic mode ofoperation [45.40]. Therefore, image lag in a-Se detec-tors is considered to be negligible. The pixel-to-pixelsensitivity variation is also negligible in a-Se detectors.The presamplingMTF of these detectors is almost equalto the theoretical MTF (sinc function) determined bythe pixel aperture [45.40].

Avalanche Gain a-SeAmorphous selenium is one of the exceptional amor-phous semiconductors, which exhibits usable impactionization, i. e., at a very high field F (above �70V=�m) holes in a-Se can gain enough energy tocreate additional EHPs through impact ionization witha useful avalanche gain of 1000 or more [45.41].Based on this hole impact ionization, a practical vac-uum device, the high-gain avalanche rushing photo-conductor (HARP) optical image sensor, has beendeveloped [45.42]. Recently, there has been an in-tense interest to replace the electron beam of HARPby a two-dimensional array of pixel electrodes utiliz-ing avalanche selenium detectors for low-dose medicalx-ray imaging [45.43]. The avalanche multiplicationmay increase the signal strength and improve the signalto noise ratio in low dose x-ray imaging applications.However, the dark current in avalanche detectors can behigh and very critical because of extremely high fieldsand the avalanche nature of the dark current. Electronsstart impact ionization at a field of � 100V=�m ina 30�m thick a-Se detector structure (it is 110V=�mfor the thickness of 10�m) [45.44]. Once electronimpact ionization starts, a self-sustaining avalanchebreakdown process is initiated and the large numberof multiplied charge carriers act to screen the externalbias. Therefore, 110V=�m field and 30�m thicknessare considered as the practical limit for an avalanchea-Se detector for x-ray imaging applications.

The potential avalanche selenium detector struc-tures for solid state flat-panel digital x-ray imagingare shown in Fig. 45.6. The prospective detector struc-tures are classified as; (a) type 1: cerium dioxide(CeO2) hole blocking and resistive interface layer (RIL)electron blocking layers, and Au bottom electrode,and (b) type 2: indium gallium zinc oxide (IGZO)hole blocking layer and Au bottom electrode. Thetop electrode for all the structures is positively biasedindium-tin-oxide (ITO). The thicknesses of the CeO2

+Bias

ITO

CeO2 RILa-Se

IGZO a-SeAu

+Bias

ITO Au

a)

b)

Fig. 45.6a,b Schematic diagrams of a-Se avalanche de-tector structures with ITO and Au top and bottom elec-trodes respectively. (a) Type 1: Cerium dioxide (CeO2)hole blocking layer and resistive interface layer (RIL) aselectron blocking layer. (b) Type 2: Indium gallium zincoxide (IGZO) hole blocking layer. (After [45.48])

and IGZO layers are � 10�30 and 375 nm respec-tively. Both CeO2 and IGZO are n-type wide bandgap(bandgap Eg D 3:3 and 3:7 eV for CeO2 and IGZO re-spectively) materials [45.42, 45]. Their bandgaps andelectron affinities are such that they both block holeinjection from electrodes but don’t block the flow ofphotogenerated electrons from the a-Se layer to the pos-itive electrode [45.45, 46]. The RIL layer is 1�m thickand it is a semi-insulating polymer, namely celluloseacetate. RIL blocks electron injection and prevents golddiffusion into the a-Se structure [45.47]. Among thereported avalanche detector structures, the type 2 struc-ture has shown the minimum steady-state dark current(1 pA=mm2 at F D 60V=�m). Another advantage ofthe type 2 structure is that it doesn’t block the flow ofphotogenerated electrons from the a-Se layer to the pos-itive electrode.

The avalanche multiplication gain g depends onthe photoconductor layer thickness L and the impactionization coefficient of the carrier. If only holes un-dergo impact ionization, the carrier multiplication isexp (˛hL), where ˛h is the impact ionization coefficientfor holes. The field dependence of the ˛h, at least overthe limited fields where avalanche is observed, can bemodeled by the following empirical relation

˛h D ˛0 exp

��F0

F

�; (45.3)

where ˛0 and F0 are the fitting parameters. Rezniket al. [45.41] carefully measured ˛h for holes in var-ious samples at room temperature and plotted as ˛h

versus 1/F. The experimental results for different sam-ples are surprisingly quite close as shown in Fig. 45.7.The symbols represent the experimental results and thesolid line represents (45.3) with ˛0 D 1:1�104 �m�1

and F0 D 1:09�103 V=�m.

PartE|45.1


0.009 0.01 0.011 0.012 0.013 0.014 0.01510–3

10–2

10–1

100

101

1/F (μm/V)

Hole impact ionization coefficient (μm–1)

8 μm a-Se layer15 μm a-Se layer35 μm a-Se layer

Fig. 45.7 The hole impact ionization coefficient as a func-tion of electric field. Symbols: experimental data. (Af-ter [45.41]). Solid line: empirical equation (45.3)

The effective mobility of holes at extremely highfields increases with increasing field. The effective holemobility at room temperature Tr is well defined and al-most independent of the preparation conditions of thesample [45.47, 49]. An empirical relation for the effec-tive hole mobility in a-Se at Tr can be expressed byfitting the experimental results, which is [45.50]

�h .F; Tr/ � 0:127C 0:745

1C exph� .F�48/

11:5

i ; (45.4)

where F is the electric field in V=�m and �h isin cm2=Vs. Figure 45.8 shows the effective holedrift mobility as a function of applied electric fieldat room temperature. The symbols and solid linerepresent the experimental results and the empiricalrelation (45.4) respectively. The experimental field-dependent effective mobility data have been extractedfrom Fig. 3 of [45.47]. The drift mobility was mea-sured in ITO/CeO2/a-Se/Sb2S3/Au and ITO/CeO2/a-Se/Sb2S3/RIL/Au detector structures. As evident fromFig. 45.8, the above empirical expression gives a closefit to the experimental results. The physical models forthe field-dependent mobility and impact ionization co-efficients are also described in the literature [45.50].

Mercuric Iodide (HgI2)Polycrystalline HgI2 (poly-HgI2) has been used asa photoconductor layer in x-ray image detectors. Ithas been prepared by either physical vapor deposition(PVD) or screen printing (SP) from a slurry of HgI2crystals using a wet particle-in-binder process [45.26].There appears to be no technological barrier topreparing large-area layers, and direct-conversion x-ray

0 20 40 60 80 100 1200

0.2

0.4

0.6

0.8

1

Electric field (V/μm)

Effective hole mobility (cm2/Vs)

With RILWithout RILInsulating

Fig. 45.8 The effective hole drift mobility as a functionof the applied electric field at room temperature. Symbols:experimental data. (After [45.47]). Solid line: empiricalequation (45.4)

AMFPI of 20� 25 cm2 (1536� 1920 pixels) and 5�5 cm2 (512� 512 pixels) size have been demonstratedusing PVD [45.26] and SP poly-HgI2 layers. The proto-type HgI2 sensors can potentially be used for radiationtherapy [45.51], mammographic [45.52], fluoroscopicor radiographic imaging. There has been active researchto improve the material properties of poly-HgI2-basedimage detectors including improving the nonuniformityby reducing the grain size [45.53]. The bandgap energyEg D 2:1 eV, the ionization energy W

˙� 5 eV and the

density of poly-HgI2 is 6:3 g=cm3. The resistivity of thismaterial is � 4�1013�cm [45.54].

The HgI2 layer is deposited onto either conductive(indium-tin-oxide-(ITO) or gold-coated) glass plates ora-Si TFT arrays. HgI2 tends to react chemically withmost metals; hence a thin blocking layer (typically,� 1�m layer of insulating polymer) is used betweenthe HgI2 layer and the pixel electrodes to prevent thereaction and also to reduce the dark current. The HgI2layer thickness varies in the range 100�500�m and thegrain size is 20�60�m. Several hundred Å of palla-dium (Pd) or Au are deposited (by direct evaporation)on top of the HgI2 layer followed by a polymer encap-sulation layer to form a bias electrode.

The dark current of HgI2 imagers increases super-linearly with the applied bias voltage [45.55]. For PVDHgI2 detectors the dark current depends strongly on theoperating temperature. It increases by a factor of abouttwo for each 6 ıC of temperature rise. It is reported thatthe dark current varies from � 2 pA=mm2 at 10 ıC to� 180 pA=mm2 at 35 ıC at F D 0:95V=�m, which isnot desirable for medical imagers (the maximum dark


E|45.1

current for medical imaging should be < 10 pA=mm2).Therefore, PVDHgI2 imagers should be operated at rel-atively low bias (preferably less than � 0:5V=�m) andrelatively low temperature (< 25 ıC). The dark currentin the SP sample is an order of magnitude smaller thanin the PVD sample and more stable against tempera-ture variation. It is possible to keep the dark currentbelow 10 pA=mm2 at temperatures up to 35 ıC and F D1V=�m [45.55]. A major disadvantage of SP detectorsis that they show 2�4 times less sensitivity comparedto PVD detectors.

Electrons have much longer range than holes inHgI2 and thus the radiation-receiving electrode is neg-atively biased to obtain a higher sensitivity. The �e�efor electrons in the SP HgI2 is in the range 10�6–10�5 cm2=V [45.26, 56], and �e�e in the PVD sampleis about an order of magnitude larger. Recently it wasreported that �e�e in PVD HgI2 is in the range 10�5–10�4 cm2=V, which is almost equal to that of single-crystalHgI2 [45.24, 57]. The reason is that the PVDHgI2layer grows in a columnar structure perpendicular to thesubstrate. Thus, a charge carrier may drift along a col-umn without having to pass through grain boundaries,where it would encounter excess trapping and/or recom-bination. Sampleswith larger grain sizesmay have fewergrain-boundary defects. Hence there is a trend of in-creasing�e�e with grain size in the PVD samples. How-ever, this trend is not observed in the SP sample, whichindicates another mechanism is responsible for its low�e�e. Larger grain sizes may cause nonuniform sensorresponse. The grain sizes must be much smaller than thepixel size to get a uniform response. �h�h in poly-HgI2is of the order � 10�7 cm2=V [45.56].

Two important drawbacks of polycrystalline sen-sors are the image lag and the pixel-to-pixel sensitivityvariation or nonuniform response. The lowest image-lag characteristics reported are � 7% first-frame lag,� 0:8% after 1 s and � 0:1% at 3 s in fluoroscopicmode (15 frames=s). The pixel-to-pixel sensitivity vari-ation reduces the dynamic range of the imagers. Therelative standard deviation of the sensitivity (standarddeviation/average value) in the latest HgI2 AMFPI is� 10%. It is reported that HgI2 image detectors withsmaller grain sizes show good sensitivity and also anacceptably uniform response. The presampling MTF ofthese detectors is almost equal to the theoretical MTF(sinc function) determined by the pixel aperture.

Cadmium Zinc Telluride (CdZnTe)CdZnTe (< 10% Zn) polycrystalline film has been usedas a photoconductor layer in x-ray AMFPI. CdZnTe iscommonly called CZT. Although CZT can be depositedon large areas, direct-conversion AMFPI of only 7:7�7:7 cm2 (512� 512 pixels) have been demonstrated us-

ing polycrystalline CZT (poly-CZT). The CZT layerthickness varies in the range 200�500�m [45.58].Temporal lag and nonuniform response were noticeablein early CZT sensors. Large and nonuniform grain sizesare believed to be responsible for the temporal lag andnonuniform response of the sensor. Recent studies showthat Cl doping makes a finer and more uniform grainstructure [45.25]. The ionization energy W

˙� 5 eV,

and the density of Cd0:95Zn0:05Te is 5:8 g=cm3. Thebandgap energy Eg of Cd0:95Zn0:05Te is 1:7 eV and theresistivity of this material is � 1011�cm [45.59].

Typically CZT is first coated onto an inert substrateand then attached to the active matrix array. The advan-tage of this technique is that the electrical propertiesand hence imaging performances of the detector canbe optimized without causing any thermal or chemicaldamage to the TFT array. The CZT film is depositedby the close-spaced sublimation method [45.60] ontoan alumina (Al2O3) substrate coated with ITO, whichforms the bias (top) electrode. A cross section of thepoly-CZT detector is shown in Fig. 45.9. The several-microns-thick ZnTe layer acts as a barrier to electroninjection and hence suppresses dark current under thenegative bias. This ZnTe layer is omitted for the positivebias. Conducting resin bumps on the pixel electrodesconnect each pixel to the CZT layer.

Introduction of Zn into the CdTe lattice increasesthe bandgap, decreases the conductivity and hence re-duces the dark current. In CZT �h decreases withincreasing Zn concentration whereas �e remains nearlyconstant. Again, addition of Zn to CdTe increases lat-tice defects and hence reduces carrier lifetimes. Thepoly-CZT has a lower density, which results in a lowerquantum efficiency � than its single-crystal counterpart.Although (for a detector of given thickness) x-ray sensi-tivity in CZT detectors is lower than in CdTe detectors,the CZT detectors show a better signal-to-noise ratioand hence give a better DQE. The measured sensitivi-ties are higher than a-Se.

The dark current of Cd0:95Zn0:05Te imagers in-creases almost linearly with F and is � 70 pA=mm2 at

Substrate (Al2O3) Bias electrode (ITO)

Barrier layer (ZnTe)

Conductive resinPixel electrode

Storage capacitor

Poly-CZT

TFT substrate

TFT

Fig. 45.9 Cross section of the polycristalline CZT detec-tor structure. (After [45.25])

PartE|45.1


F D 0:25V=�m [45.27], which makes it unsuitable forapplications requiring long exposure times. The darkcurrent would be expected to decrease with increas-ing Zn concentration due to the increased Eg. This hasin fact been demonstrated [45.25], i. e., the dark cur-rent in Cd0:92Zn0:08Te sensors is 40 pA=mm2 at F D0:4V=�m.

The mobility lifetime products of both electrons andholes in poly-CZT are less than that in single-crystalCZT. The values of �� in single-crystal Cd0:9Zn0:1Teare in the range 10�4�10�3 cm2=V (electrons) and10�6�10�5 cm2=V (holes) [45.61]. But the values of�� in poly-Cd0:95Zn0:05Te are � 2�10�4 cm2=V (elec-trons), and � 3�10�6 cm2=V (holes) [45.62, 63]. Since�e�e � �h�h in CZT, negative bias to the radiation-receiving electrode is the preferred choice for bettersensitivity and temporal response.

The relative standard deviation of the sensitivity(standard deviation/average value) in the latest CZTAMFPI is � 20% [45.25]. The image-lag characteris-tics reported are � 70% first-frame lag, � 20% afterthree frames and 10% at 1 s in fluoroscopic mode(30 frames=s). In single-pulse radiographic mode thefirst-frame lag is less than 10% [45.25]. The longerimage-lag characteristics of CZT sensors in fluoro-scopic mode imply that it is not yet suitable for fluo-roscopic applications. The presampling MTF of CZTdetector is � 0:3.150�m pixel size) at the Nyquist fre-quency fN (theoretical MTF, sinc function, is � 0:64 atfN D 3:3 line pairs per mm), where the MTF of CsI im-agers is less than 0:2 [45.27].

Lead Iodide (PbI2)Polycrystalline PbI2 photoconductive layers have beenprepared using PVD at a substrate temperature of 200–230 ıC. A deposition of several hundred Å of palla-dium (Pd) is used to form the top electrode. Grainsare described as hexagonal platelets with the longestdimensions being 10�m or less. The platelets growperpendicular to the substrate, producing films thatare less dense (3�5 g=cm3) than bulk crystalline ma-terial (6:2 g=cm3). There appears to be no technolog-ical barrier to preparing large-area layers, and direct-conversion AMFPI of 20� 25 cm2 size (1536� 1920pixels) have been demonstrated [45.64]. Coating thick-ness varies in the range 60�250�m and prototypePbI2 imagers have been used for radiographic imag-ing [45.64]. The bandgap energy Eg D 2:3 eV, and theionization energy W

˙� 5 eV. The resistivity of this

material is in the range 1011�1012�cm.Lead iodide detectors have a very long image-lag

decay time. The image lag depends on the exposurehistory. The image-lag characteristics reported are �75% first-frame lag, � 15% after 3 s in fluoroscopic

mode (15 frames=s), whereas in single-pulse radio-graphic mode the first-frame lag is less than 50% anddrops below 1% within 1 s [45.64]. The long image-lagcharacteristics of PbI2 in fluoroscopic mode imply thatit is unsuitable for fluoroscopic applications.

The dark current of PbI2 imagers increases sublin-early with the applied bias voltage. The dark current isin the range 10�50 pA=mm2 at F D 0:5V=�m [45.64]and 0:1�0:45 nA=mm2 at F D 1:0V=�m [45.65],much higher than PVD HgI2 detectors, making it un-suitable for long-exposure-time applications. The pre-sampling MTF is � 0:35 (127�m pixel size) at fN(theoretical MTF, sinc function, is � 0:64 at fN D3:93 lp=mm), where theMTF of CsI imagers is less than0:2 [45.64]. The resolution of PbI2 imagers is accept-able but slightly worse than that of HgI2 imagers. Also,the x-ray sensitivity of PbI2 imagers is lower than thatof HgI2 imagers. The pixel-to-pixel sensitivity variationin PbI2 imagers is substantially lower.

The �� product of holes and electrons in PVDPbI2 are 1:8�10�6 cm2=V and 7�10�8 cm2=V re-spectively [45.64]. �h in poly-PbI2 is in the range0:02�0:15 cm2=Vs whereas �h in single-crystal PbI2 is2 cm2=Vs [45.66]. This indicates that �h in poly-PbI2 iscontrolled by shallow traps, probably introduced at thegrain boundaries.

Lead Oxide (PbO)The large-area deposition requirement is compatiblewith the use of polycrystalline PbO (poly-PbO) film asa photoconductor layer in AMFPI. Direct-conversionflat-panel x-ray imagers of 18� 20 cm2 (1080� 960pixels) from a poly-PbO with film thicknesses of �300�m have been demonstrated [45.28]. One advan-tage of PbO over other x-ray photoconductors is theabsence of heavy-element K-edges for the entire di-agnostic energy range up to 88 keV, which suppressesadditional noise and blurring due to the K-fluorescence.The ionization energy W

˙� 8 eV and the density of

crystalline PbO is 9:6 g=cm3 whereas its density is re-duced by almost 50% in poly-PbO. The bandgap energyEg of PbO is 1:9 eV and the resistivity of this material isin the range (7�10) �1012�cm [45.28].

Lead oxide photoconductive polycrystalline layershave been prepared by thermal evaporation in a vac-uum chamber at a substrate temperature of � 100 ıC.An evaporated layer of Al, Au or Pd was used to forma top electrode of thickness 100�200 nm [45.67]. ThePbO layer consists of very thin platelets of a few mi-crons thickness and has a density of � 50% of thesingle-crystal density. PbO slowly degrades if it is ex-posed to air under normal ambient temperature buta few hours exposure is acceptable. However, in thelong term PbO reacts with water and CO2 causing


E|45.1

an increase in dark current and a decrease in x-raysensitivity. Therefore, a polymer or a semiconductor(e.g., doped a-Se) passivation layer is used to pre-vent exposure to the atmosphere. This is usually placedbetween the photoconductor layer and the top metalelectrode. The passivation layer avoids degradation ofthe PbO layer and reduces the dark current [45.67].The dark current in PbO sensors is � 40 pA=mm2 atF D 3V=�m [45.28]. The �e�e and �h�h in poly-PbOare 3:5�10�7 and � 10�8 cm2=V respectively [45.68].The lag signal in fluoroscopic mode is in the range3�8% after 1 s. The presampling MTF of a PbO detec-tor is � 0:5 (184�m pixel size) at fN (theoretical MTF,sinc function, is � 0:64 at fN D 2:72 lp=mm) [45.28].

Thallium Bromide (TlBr)Polycrystalline thallium bromide (poly-TlBr) has notyet been used in an AMFPI but has been used in a large-area (9 in-diameter) direct-conversion detector calledan x-ray-sensitive electron-beam image tube (XEBIT).The operational principle of the XEBIT is similar tothe standard light-sensitive vidicon that was utilizedextensively in the commercial television industry. TheXEBIT can replace an x-ray image intensifier coupledto a video camera using relay lenses with a singledirect-conversion device [45.69]. The typical TlBr layerthickness is 300�m.

The ionization energyW˙

� 6:5 eV and the relativedielectric constant of TlBr is 31 [45.70]. The bandgapenergy Eg of TlBr is 2:7 eV and its resistivity is �5�109�cm under ambient conditions. At room tem-perature the dominant contribution to the dark currentis ionic conductivity [45.69]. The ionic conductivityhas an exponential dependence on temperature; the con-ductivity decreases by an order of magnitude for every19 ıC temperature decrease. Therefore, the dark currentcan be greatly decreased by Peltier cooling. �h�h and�e�e in TlBr are � 1:5�10�6 and � 4�10�7 cm2=Vrespectively [45.69, 70].

Organic PerovskitesYakunin et al. [45.71] reported that a thick layer (�100�m) of polycrystalline methylammonium lead io-dide (poly-MAPbI3 where MA is CH3NH3) perovskite(grain size is larger than 0:25�m) can be uniformlydeposited over large areas using a solution-based syn-thesis technique without affecting the underlying AMAelectronics. This material also shows a reasonable x-rayabsorption coefficient and its x-ray sensitivity is compa-rable to a-Se [45.71]. Therefore, MAPbI3 can be a po-tential candidate for large-area AMFPI and thus opensthe door to a new class of perovskite large-area x-raysensors. MAPbI3 is a direct bandgap semiconductorsof bandgap energy of 1:6 eV and the relative dielec-

tric constant of 28. Its density is 4:3 g=cm3 [45.72].The electron and hole mobilities in polycrystalline filmsare � 8 and � 15 cm2=Vs respectively [45.71]. Thedark resistivity is � 109 �cm. The effective masses ofelectrons and holes are 0:23 m0 and 0:29 m0 respec-tively [45.73], where m0 is the free electron mass.

Although the reported 60�m-thick detector struc-ture [45.71] shows a reasonable level of photocurrentfor the x-ray fluence of 1:4�107 photons=mm2s, thedark current is almost half of the photocurrent. More-over, the dark current increases linearly with the appliedfield and the detectors shows a photoconductive gain,which indicates that the contacts are ohmic-type. Asa result, the dark current could be unacceptably largeat high applied field (the high field is needed to achieveacceptable charge collection efficiency and x-ray sen-sitivity). Though the ohmic contact is required forthe photovoltaics, the x-ray detectors need blockingcontacts in order to minimize the noise and enhancethe DQE. Therefore, the optimization of the structure(for example, the appropriate metal contacts and theblocking layers) is essential for their use in practicalAMFPIs.

45.1.3 Summary and the Future

The material and imaging properties of potential photo-conductors for x-ray image detectors are summarizedin Tables 45.1 and 45.2. Stabilized a-Se is currentlythe best choice of photoconductor for mammographicx-ray image detectors. The next closest competitor forboth mammography and general radiology is the poly-HgI2 imagers, which show excellent sensitivity, goodresolution, and acceptable dark current, homogeneityand lag characteristics. However, the long-term sta-bility of HgI2 imagers has not been as thoroughlystudied as stabilized a-Se detectors. Both the dark cur-rent and the image-lag characteristics of CZT, PbI2and PbO detectors are worse than those of HgI2 de-tectors. However, the x-ray detectors made with CZTphotoconductive layers should be mechanically andchemically more stable than HgI2-based detectors. Themain drawback of a-Se detectors are their low conver-sion gain, which particularly affects the imaging sensorperformance at low exposure. This can be overcomeby utilizing the avalanche multiplication technique inthe a-Se layer [45.74] and/or using on-pixel amplifi-cation [45.2]. However, further research is necessaryto demonstrate basic operation as well as to examinethe long-term stability of detectors utilizing these tech-niques. The main drawbacks of polycrystalline sensorsare the image lag and the nonuniform response. Makingsmaller, finer and more uniform grain size in polycrys-talline sensors may overcome these drawbacks.

PartE|45.2


Table 45.1 Material properties

Photoconductor,state and preparation

Eg (eV) W˙

(eV)Density(g=cm3)

Resistivity(�cm)

Electrons�e(cm2=Vs),�e�e(cm2=V)

Holes�h(cm2=Vs)�h�h(cm2=V)

Stabilized a-Se, vacuum deposition 2:1�2:2 � 45 at10V=�m

4:3 1014�1015 �e D0:003�0:006�e�e D0:3�10�6�10�5

�h D 0:12�h�h D10�6�6�10�5

HgI2, polycrystalline, PVD 2:1 5 6:3 � 4�1013 �e D 88�e�e �10�5�10�4

�h D 3�4�h�h � 10�6

HgI2, polycrystalline, SP 2:1 5 6:3 � 4�1013 �e�e �10�6�10�5

�h�h � 10�7

Cd0:95Zn0:05Te, polycrystalline,vacuum deposition

1:7 5 5:8 � 1011 �e�e � 2�10�4 �h�h � 3�10�6

PbI2, polycrystalline, PVD 2:3 5 3�5 1011�1012 �e�e D 7�10�8 �h D 0:02�0:15,�h�h � 2�10�6

PbO, polycrystalline, vacuum depo-sition

1:9 8�20 4:8 7�10�1012 �e�e � 3:5�10�7

�h�h � 10�8

TlBr, polycrystalline 2:7 6:5 7:5 � 5�109 at20 ıC

�e�e � 4�10�7 �h�h � 1:5�10�6

MAPbI3, polycrystalline 1:6 5 4:3 � 109 �e D 8 �h D 15

Table 45.2 Imaging properties

Photoconductor,state and preparation

Typicaloperating F(V=�m)

Dark current.pA=mm2/

Lag (fluoroscopicmode of operation)

Uniformity/sensitivity variation(standard deviation/average value)

Stabilized a-Se, single layer � 10 < 10 up toF D 20V=�m

< 2% after 33ms Negligible

Stabilized a-Se, multilayer (pinor nip)

� 10 < 1 up toF D 20V=�m

< 2% after 33ms Negligible

HgI2, polycrystalline, PVD � 0:5 � 6 atF D 0:5V=�m

� 7% after 66ms � 10%

HgI2, polycrystalline, SP � 1:0 � 8 atF D 1:0V=�m

� 7% after 66ms � 10%

Cd0:95Zn0:05Te, polycrystalline � 0:25 � 25 atF D 0:25V=�m

� 70% after 33ms � 20%

PbI2, polycrystalline, PVD � 0:5 10�50 atF D 0:5V=�m

� 75% after 66ms ?

PbO, polycrystalline � 1:0 40 at F D 3V=�m field 3�8% after 1 s ?TlBr, polycrystalline � 0:5 ? ? ?

45.2 Dark Current Limitations

As mentioned earlier, a large electric field (e.g., �10V=�m in conventional a-Se detectors) is appliedacross the x-ray detectors in order to collect photo-generated EHPs [45.43]. One of the most importantattributes of a photoconductor is the dark current, whichshould be negligibly small, since it is a source of noiseand thus reduces the detectivity and dynamic range ofthe detector. The dark current in detectors is usuallyvery high right after applying the bias voltage and de-

cays with time. In most of the cases, it reaches a plateauwithin the time range of 100�10 000 s [45.36, 37, 47].

The dark current in x-ray detectors may have threeorigins:

1. The depletion of carriers from the defect stateswithin the bandgap after applying the bias

2. Steady-state thermal generation of carriers in thebulk

Photoconductors for X-Ray Image Detectors 45.2 Dark Current Limitations 1137Part

E|45.2

3. Carrier injections from the metal contacts towardsthe photoconductor layer.

45.2.1 Charge Carrier Depletion

The depletion of the majority carrier from the bulk andinterface defect states within the mobility gap can con-stitute a transient dark current decay behavior. Afterapplying the bias, majority carriers are depleted andthe steady-state quasi-Fermi level EFD becomes slightlydifferent from the Fermi level EF. Let us consider a-Seas an example. a-Se is slightly p-type and EF in a-Se atzero bias is slightly below the midgap. After applyingthe bias, holes are depleted and the steady-state quasi-Fermi level EFD lies above EF as shown in Fig. 45.10.The temporal behavior of the carrier depletion processis determined by the detrapping time constants. Thetime-dependent hole depletion rate due to carrier de-trapping from the bulk a-Se is [45.48, 75]

gd.t/ DEcZ

Ev

N.E/

�r.E/ŒfFD.E/� fF.E/� exp

�� t

�r.E/

dE ;

(45.5)

where

fF.E/ D 1

1C exph.E�EF/kBT

i ; (45.6)

fFD.E/ D 1

1C exph.E�EFD/

kBT

i : (45.7)

The mean detrapping or mean release time constant is

�r.E/ D ��1 exp

2

4

E �EV �ˇpf

pF�

kBT

3

5 ; (45.8)

where N.E/ is the density of states (DOS) of the photo-conductor at energy E in the midgap, � is the attempt-to-escape frequency, kB is the Boltzmann constant, T isthe absolute temperature, t is the instantaneous time (inseconds), ˇpf D p

e3=�"s is the Poole–Frenkel coeffi-cient, e is the elementary charge, "s is the permittivity ofthe photoconductor, and EC and EV are the conductionand valence band edges respectively. Equation (45.5)represents the rate of hole depletion per unit volume,which decays with time.

The depleted holes drift under the influence ofthe electric field and induce a current in the detector.Currents resulting from the drifting of carriers in the

Density of states, N(E)

EC

EF

EFD

EV

Energy

Valence band

Conduction band

Fig. 45.10 A schematicrepresenting the car-rier depletion fromthe mobility gap inslightly p-type amor-phous/polycrystallinematerials after applyingthe bias. The shaded arearepresents the approxi-mate amount of depletedcarriers. The figure showsan arbitrary DOS foran amorphous material.(After [45.48])

photoconductive detectors are due entirely to induc-tion, which can conveniently be calculated by usingthe Shockley–Ramo theorem [45.76]. The transient cur-rent density due to the hole depletion from the bulkis [45.77],

Jdpb.t/ D eL

2gd.t/ : (45.9)

Similarly, the time-dependent hole depletion currentdensity due to carrier detrapping from the interface(e.g., the top interface at positive bias to the top elec-trode) is [45.48],

Jdpi.t/ D e

EcZ

Ev

Di.E/

�r.E/ŒfFD.E/�fF.E/� exp

�� t

�r.E/

dE ;

(45.10)

where Di.E/ is the density of interface states (cm�2

eV�1). Note that (45.5)–(45.10) can also applied toslightly n-type photoconductors (e.g., HgI2) by inter-changing fF and fFD [45.78].

45.2.2 Steady-State Thermal Generation

The defect states close to the middle of the bandgapof the x-ray photoconductor have a significant proba-bility for thermal excitation of both types of carriers.The bandgap/mobility gap of x-ray photoconductors isquite large and thus the band-to-band thermal carriergeneration is expected to be negligible. Therefore, thesteady-state thermal generation rate is dominated bythe emission from traps within kBT of EFD as shownin Fig. 45.11. If the excitation rates for electrons andholes are equal, EFD is very close to the middle of thebandgap. The generation rate for a fully depleted sam-ple is determined by the average carrier release time and

PartE|45.2


Density of states, N(E)

EC

EFD

EV

Energy

Valence band

Conduction band

kBT

Thermal generation

Fig. 45.11 A Schematicrepresenting the thermalgeneration of carriers fromthe mobility gap of amor-phous/polycrystallinematerials near thequasi-Fermi level EFD.(After [45.48])

Photoconductor

EC

Metal

EFm

TAT

Energy

Schottky BL

E

φe

Fig. 45.12 A schematicrepresenting the electroninjection process fromthe metal towards thephotoconductor layer.TAT and BL stand forthe thermally assistedtunneling and barrierlowering respectively.(After [45.48])

can be written as [45.48],

g D N.EFD/kBT� exp

2

4�EC �EFD �ˇpf

pF�

kBT

3

5 ;

(45.11)

where N.EFD/ is the density of states at energy EFD inthe midgap. It is assumed in (45.11) that the density ofstates is constant over kBT near EFD. Assuming that theliberated carriers are not lost by trapping or recombina-tion (any loss of carriers can be reflected in the effectivevalue of g), the steady-state thermal generation currentis

Jth.t/ D egL : (45.12)

45.2.3 Carrier Injection

The metal/photoconductor contacts in detectors aregenerally blocking in nature and the carrier injection ismainly controlled by the Schottky emission. Once thecarriers are injected into the photoconductive layer, they

move by the drift mechanisms (the diffusion componentof current is negligible compared to its drift compo-nent because of the very high applied bias) [45.79].Therefore, the injected current densities through themetal/photoconductor contacts are [45.79],

Je.t/ D eNCvde.t/

�vr

vde C vr

�

� exp

��e � �e.F/

kBT

(45.13)

and

Jh.t/ D eNVvdh.t/

�vr

vdh C vr

�

� exp

��h � �h.F/

kBT

; (45.14)

where vr D A�T2=eNC is the thermal velocity, A� isthe effective Richardson constant, NV.C/ is the effec-tive density of states in the valence (conduction) band,vd.t/ � �F is the instantaneous drift velocity of the in-jected carriers, � is the effective mobility of carriers,and � is the effective barrier height for injecting carriersfrom the metal into the photoconductor layer, consider-ing the effect of surface states. The subscripts e and hstand for electrons and holes respectively. For the po-tential x-ray photoconductors, vr � vdh or vde, the terminside the square bracket of (45.13) and (45.14) be-comes unity, and the injection current is dominated byvdh or vde depending on the particular photoconductor(e.g., vdh � vde in a-Se) [45.79].

The quantity �.F/ within the exponentialin (45.13) and (45.14) is the field-dependent barrierlowering due to other effects such as charge rearrange-ment in the interfacial layer, the image force barrierlowering, and/or thermally assisted tunneling as shownin Fig. 45.12. For the metal-to-semiconductor carrier in-jection, the term �.F/ can be written as

�.F/ Ds

e3F

4�"sC e˛F2 : (45.15)

The first term in (45.15) is the image force barrier low-ering (Schottky barrier lowering), and the second termis the optimum barrier lowering due to the thermally as-sisted tunneling. The potential barrier for the injectingcarriers from the metal contact under a strong appliedfield is essentially a triangular barrier. Note that thethermally activated tunneling is much more probablethan the direct tunneling. In thermally activated tun-neling process, as the vibration energy Ep increases,the tunneling barrier decreases and hence the proba-bility increases. On the other hand the population of

Photoconductors for X-Ray Image Detectors 45.3 Metrics of Detector Performance 1139Part

E|45.3

the energy level Ep decreases with increasing Ep pro-portionally to exp.�Ep=kBT/. Thus, there exists anoptimum energy level from where the resultant escapeprobability is the maximum, which is proportional toexp.˛F2=kBT/ [45.80]. The expression of ˛ is deter-mined by considering the tunneling probability througha triangular barrier together with Boltzmann occupationprobability and given by [45.80]

˛ D e¯2

24m� .kBT/2; (45.16)

where h is the modified Plank constant and m� is theeffective mass of the photoconductor.

Concluding Remarks on Dark CurrentOne of the advantages of a-Se over competinglarge-area x-ray photoconductors is the compara-tively smaller dark current under typical operatingfields [45.36, 81]. The operating field refers to the ap-plied field that ensures the x-ray generated chargesare collected; i. e., the sensitivity is not limited bythe charge collection efficiency [45.82]. For example,the dark current that has been recently reported forPbI2 [45.81] is in the range 0:1�0:45 nA=mm2 at a fieldof 1V=�m. Under normal operation, the field needsto be greater than 1V=�m to ensure good charge col-lection but this would also mean a dark current that isgreater than 1 nA=mm2, which is much higher than thedark current in typical a-Se detectors even at the oper-ating field of 10V=�m [45.36].

As mentioned earlier, the conventional a-Se detec-tor can be either n-i-p or n-i type. The experimentalresults in the literature show that the main source ofthe dark current is the injection of holes from the metalcontacts and the bulk thermal generation current is neg-ligible [45.36, 37]. The above results are consistent withthe facts that the bulk thermal generation and elec-tron injection currents are negligible compared to thehole injection current because of a large bandgap and

a very low electron mobility in a-Se respectively. Freyet al. [45.36] have also shown that the dark currenthas a strong dependence on the n-like layer thickness;it decreases abruptly with increasing the n-like layerthickness (� 4 to 8�m). The carrier injection, fromthe metal to the semiconductor, depends on the internalelectric field at the metal/a-Se interface. It is believedthat the electric field right after applying the bias isalmost uniform but quickly decreases at the interfacedue to high initial current and high carrier trappingin the blocking layers, which reduces carrier injectionand hence reduces the dark current. The typical elec-tric field profiles, after applying the bias on the n-i-pand n-i structures, are illustrated in Fig. 45.5. To re-duce the contact electric field and dark current, then-like layer must have a sufficient amount of trap cen-ters and the energy depths of these trap centers shouldbe � 0:75�0:8 eV or more from the valence band mo-bility edge. The shallower trap levels are unable toretain a sufficient amount of trapped charge and thedeeper trap centers create longer transient times to reacha steady state dark current. The detailed descriptions ofthe dark current mechanisms are given in [45.48].

At the normal operating fields (less than 10V=�m)of conventional a-Se or poly-HgI2 x-ray detectors,the thermal generation current is negligible as com-pared to the injection currents from the metallic elec-trodes [45.48, 78]. However, the thermal generationrate can be increased exponentially with field becauseof Poole–Frenkel or thermally assisted tunneling ef-fects [45.83]. Therefore, both the injection and thermalgeneration currents should be considered to determinethe dark current in a-Se avalanche detectors. The ex-cessive dark current has been one of the factors thatlimits the highest operating field. An acceptable level ofdark current up to an electric field as high as 60V=�min some a-Se detector structures has recently been re-ported [45.45]. The injection current is minimized andthe dark is slightly higher than the thermal generationcurrent in the type 2 structure in Fig. 45.6.

45.3 Metrics of Detector Performance

X-ray sensitivity, resolution in terms of modulationtransfer function (MTF), detective quantum efficiency(DQE), image lag and ghosting are often consideredas the metrics of imaging performance. For most prac-tical applications, the spatial-frequency-dependent (f -dependent) detective quantum efficiency, DQE(f ), isthe appropriate metric of overall system performanceand is unity at all f for an ideal detector. The detectorperformance depends critically on the photoconductormaterial properties such as the mobility, carrier trap-

ping (both shallow and deep), EHP creation energy,x-ray attenuation and absorption coefficients. The ma-terial properties such as carrier mobility, EHP creationenergy, x-ray attenuation and absorption coefficients ina well-defined photoconductor are almost constant, butthe carrier lifetimes may vary from sample to sample.Shallow and deep trapping are particularly responsiblefor image lag and ghosting respectively. The effects ofcharge-transport properties (��) and the attenuation co-efficient of the photoconductor material on the detector

PartE|45.3


performance depends on L and F through the followingnormalized parameters

� D normalized attenuation depth

(attenuation depth/thickness)

D 1

.˛L/;

xe D normalized electron schubweg

(electron schubweg per unit thickness)

D �e�eF

L; and ;

xh D normalized hole schubweg

(hole schubweg per unit thickness)

D �h�hF

L;

where ˛ is the linear attenuation coefficient of the pho-toconductor, �e.h/ is the mobility and �e.h/ is the deep-trapping time (lifetime) of electrons (holes). Equiva-lently, xe and xh are the normalized carrier lifetimes(carrier lifetimes per unit transit time) for electrons andholes respectively. The ranges of these normalized pa-rameters for the three most promising photoconductors(a-Se, poly-HgI2 and poly-CZT) for use in x-ray imagedetectors are given in Table 45.3. The combined effectsof charge-transport properties (mobility and carrier life-time), operating conditions (F and E), photoconductorthickness, and the attenuation coefficient of the photo-conductormaterial on the imaging characteristics (x-raysensitivity, DQE, MTF and ghosting) are examined inthe following sections. It must be emphasized that thephotoconductor thickness L and the operating field Fare as important to the overall performance of the de-tector as the material properties of the photoconductoritself, a point that will become apparent in the resultspresented in this chapter.

45.3.1 X-Ray Sensitivity

The x-ray sensitivity (S) of a photoconductive detectoris defined as the collected charge per unit area per unitexposure of radiation and is considered an importantperformance measure for a superior image. High S per-

Table 45.3 The values of �, xe and xh for a-Se, poly-HgI2 and poly-CZT detectors

Photoconductor �e�e (cm2=V)

�h�h (cm2=V)

F(V=�m)

E (keV) L (mm) xe xh

Stabilized a-Se �e�e D 0:3�10�6�10�5 � 10 20 0:2 1:5�50 5�300 0:24�h�h D 10�6�6�10�5 60 1:0 0:3�10 1�60 0:98

Poly-HgI2 �e�e � 10�6�10�5 0:5�1 20 0:15 0:7�7 � 0:1 0:21�h�h � 10�7 60 0:3 0:35�3:5 � 0:05 0:85

Poly-Cd0:95Zn0:05Te �e�e � 2�10�4 � 0:25 20 0:3 � 17 � 0:25 0:26�h�h � 3�10�6 60 0:3 � 17 � 0:25 0:89

mits the use of low detector radiation-exposure levels,which also increases the dynamic range of the AMPFI.The selection of the x-ray photoconductor is highly in-fluenced by the value of S.

The value of S can be considered to arise in termsof three controlling factors:

1. The amount of radiation actually attenuated fromthe incident radiation that is useful for the gen-eration of electron–hole pairs (EHPs), which ischaracterized by the quantum efficiency � of thedetector and depends on the value of ˛ (the linearattenuation coefficient) of the photoconductor and Lthrough � D 1�e�˛L, where the value of ˛ is x-rayphoton-energy-dependent.

2. The generation of EHPs by x-ray interactions,which is characterized by the value of W

˙of the

photoconductor and the average absorbed energyEab per attenuated x-ray photon of energy E, whereW

˙depends on the material properties of the pho-

toconductor, and Eab depends on the incident x-rayphoton energy [45.84] and the material properties.Note that Eab D E.˛en=˛/, where ˛en is the energyabsorption coefficient of the photoconductor, whichdepends on the x-ray photon energy E.

3. How much of the x-ray generated charge is actuallycollected in the external circuit. This is character-ized by the charge-carrier drift mobilities (�) andlifetimes (�), of the applied F and L.

The S of an x-ray image detector can be normal-ized with respect to the maximum sensitivity (S0) thatwould arise if all the incident radiation were absorbedand all the liberated carriers were collected. Neglect-ing secondary photon interactions, the expression for S0is [45.32]

S0 D0

@5:45�1013 e˛air�air

�W

˙

1

A˛en

˛

�; (45.17)

where e is the elementary charge, while ˛air and �air arethe energy absorption coefficients of air and its density.If W

˙is expressed in eV, ˛air=�air is in cm2=g and ex-

posure in (45.17) is in R, then S0 is in units of C=cm2R.Thus S0 is a constant that depends on the x-ray photon


E|45.3

energy E and the material properties of the photocon-ductor, sinceW

˙is a material property that can usually

be taken as constant for a given material. For thosematerials (e.g., a-Se) that have a significant F- and/orE-dependentW

˙, then S0 depends on F and/or E.

The quantity s D S=S0 takes into account the x-rayabsorption and charge-transport effects and is called thecharge-collection and absorption-limited normalizedsensitivity. It should be emphasized that s is a quan-tity that is determined by the x-ray absorption profile,photoconductor thickness and the charge-collection ef-ficiency. The s of an x-ray detector considering smallsignal operation, a constant � and a single deep-trapping time (lifetime) � for each type of carrier (holesand electrons) and neglecting carrier diffusion is givenby [45.82],

S

S0D xh

"1� e�

1�

�C 1

�xh

� 1

e�

1xh � e�

1�

�#

C xe

"1� e�

1�

�� 1

�xeC1

1� e�

1�

�1xe

�#

D sh.xh; �/C se.xe; �/ D s.xh; xe; �/ ;

(45.18)

where subscripts h and e refer to holes and electronsrespectively.

The two square brackets on the right-hand side ofthe normalized sensitivity s expression (45.18) rep-resent the relative contributions of hole and electrontransport to the overall sensitivity for a given �. It isassumed in (45.18) that the radiation-receiving side ofthe detector is biased positively. If the bias polarity isreversed, then xe and xh must be interchanged. The ex-pression in (45.18) applies for incident radiation that ismonoenergetic and has to be appropriately integratedover the radiation spectrum of a practical polyenergicx-ray source by considering the x-ray photon-energy-dependent terms W

˙, ˛ and ˛en. Equation (45.18)

applies to an isolated photoconductor sandwiched be-tween two large-area parallel-plate electrodes (smallpixel effects are excluded) [45.85] and operating un-

Table 45.4 X-ray sensitivity of a-Se, poly-HgI2 and poly-CZT detectors using the normalized parameters from Ta-ble 45.3 (E is the x-ray photon energy)

Photoconductor E S0 s D S=S0 S(�C=cm2R)

(keV) (�C=cm2R) Positive bias Negative bias Positive bias Negative biasStabilized a-Se 20

600:2445:37

0:9�0:980:39�0:64

0:8�0:980:35�0:62

0:22�0:242:1�3:38

0:2�0:241:88�3:35

Poly-HgI2 2060

2:7538:54

0:25�0:290:15�0:3

0:53�0:810:21�0:4

0:7�0:816:76�11:21

1:46�2:248:18�15:6

Poly-Cd0:95Zn0:05Te

2060

335:87

� 0:456� 0:41

� 0:85� 0:51

� 1:37� 14:69

� 2:55� 18:12

der a constant F (small signal case). An excellent fitof (45.18) to experimental data on poly-HgI2 is givenin [45.56].

The value of s is always less than unity since S fora photoconductor of finite thickness in which carriercollection is not perfect is always less than S0. Note thats.xh; xe; �/ D sh C se D 1 when all the incident radia-tion is absorbed and all the charges are collected, i. e.,xh, xe � 1 and� � 1. The sensitivity is then simply S0and is controlled byW

˙.

The sensitivity is mainly controlled by the chargesthat have the same polarity as the bias on the radiation-receiving electrode: holes for positive bias and electronsfor negative bias [45.55, 81]. The extent of the dispar-ity between sh and se depends on �. The disparity isstronger for lower �, which can be understood by not-ing that electron and hole generation do not occur uni-formly throughout the thickness of the sample but ratherit is more effective closer to the radiation-receivingelectrode. An advantage of a-Se detectors is that bothxe and xh are much greater than one, which is not thecase for other photoconductors, as shown in Table 45.3.Therefore, charge collection in a-Se detectors is closeto unity and the normalized sensitivity is controlled bythe quantum efficiency of the detector. The S and s ofdifferent photoconductive detectors using the normal-ized parameters from Table 45.3 are given in Table 45.4.The values ofW

˙for a-Se were taken from the work of

Blevis et al. [45.30]. At F D 10V=�m,W˙

D 42:5 and46 eV for x-ray photon energies of 60 and 20 keV re-spectively. The maximum x-ray sensitivity S0 of chestradiographic detectors (E D 60 eV) is much higher thanof mammographic detectors (E D 20 eV) because of thelower ˛air=�air values in (45.17) at higher x-ray photonenergies [45.84]. The value of S0 in a-Se detectors ismuch lower than in poly-HgI2 and poly-CZT detectorsbecause of the higher value ofW

˙in a-Se.

45.3.2 Detective Quantum Efficiency

DQE measures the ability of the detector to transfer sig-nal relative to noise from its input to its output. Therandom nature of image quanta gives rise to random

PartE|45.3


fluctuations in image signals, which contributes to im-age formation and hence creates random noises. Thescattering of image quanta gives rise to image blurring,which is quantified by the MTF(f ). Images are partiallydegraded by various sources of statistical fluctuationsthat arise along the imaging chain. The relative increasein image noise due to an imaging system as a func-tion of spatial frequency f is expressed quantitativelyby DQE(f ) which represents the signal-to-noise trans-fer efficiency for different frequencies of information inan image. DQE(f ) is defined as

DQE.f / D SNR2out.f /

SNR2in.f /

; (45.19)

where SNRin and SNRout are the signal-to-noise ratios(SNR) at the input and output stages of the image detec-tor respectively. DQE(f ) is unity for an ideal detector.For simplicity, we are often interested in measuringDQE(f D 0) of an imaging detector since it representsthe signal quality degradation due to the signal andnoise transfer characteristics of the system without con-sidering signal spreading.

The random nature of charge-carrier trapping(i. e., incomplete charge collection) in the photocon-ductor layer creates fluctuations in the collected chargeand hence creates additional noise. Thus carrier trap-ping degrades signal-to-noise performance of the im-age and reduces the DQE. Kabir and Kasap [45.86]have examined the effects of charge-carrier trappingon the DQE(0) of an a-Se detector by consideringdepth-dependent conversion gain and depth-dependentcharge-collection efficiency in the cascaded-linear-system model [45.87, 88]. Below we apply the DQE(0)model in [45.86] to potential photoconductive detectorssuch as a-Se, poly-HgI2 and poly-Cd0:95Zn0:05Te detec-tors for fluoroscopic applications to study and comparetheir DQE(0) performance.

Figure 45.13 shows DQE(0) as a function ofx-ray exposure for a-Se, HgI2, and CZT detectors fora 60 keV x-ray beam. The x-ray exposure (X) is var-ied from 0:1�R to 10�R, which is the range ofx-ray exposure for fluoroscopic applications. We as-sume that the pixel area, A D 200� 200�m, and theeffective fill factor is 1:0 for all the photoconduc-tors. The average E is 60 keV and the additive elec-tronic noise (Ne) is assumed to be 2000 electronsper pixel. The following transport and operating pa-rameters are used in Fig. 45.13: for a-Se detectors,L D 1000�m, F D 10V=�m, W

˙� 43 eV, �e�e D

10�6 cm2=V and �h�h D 10�5 cm2=V; for HgI2 detec-tors, L D 260�m, F D 0:5V=�m, W

˙D 5 eV, �e�e D

6�10�6 cm2=V and �h�h D 10�7 cm2=V; and for CZTdetectors, L D 270�m, F D 0:25V=�m, W

˙D 5 eV,

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.11086420

DQE(0)

Exposure (μR)

Quantum efficiencyCZT

a-Se HgI2

Fig. 45.13 DQE(0) versus x-ray exposure for a-Se, poly-HgI2, and poly-CZT detectors and for a 60 keV monoener-getic x-ray beam. The electronic noise is 2000 e=pixel. It isassumed that F D 10V=�m for a-Se, 0:5V=�m for HgI2and 0:25 V=�m for CZT

�e�e D 2�10�4 cm2=V and �h�h D 3�10�6 cm2=V.The radiation-receiving electrode is biased positivelyfor a-Se detectors and negatively for HgI2 and CZTdetectors. � � 0:98 is for all photoconductors. � D1 � exp .�1=�/ � 0:64 is the maximum achievableDQE(0) if all the liberated charges are collected.

The DQE for the CZT detector is relatively un-changed over the whole exposure range due to a largeconversion gain (low W

˙) and high charge-collection

efficiency (good transport properties). The DQE for a-Se detectors is small at low exposures because of itsrelatively low conversion gain and is controlled by theadded electronic noise. As the x-ray exposure increases,each pixel receives more photons. The relative contri-bution of electronic noise to the total noise becomesless important and the DQE increases. The DQE forHgI2 detectors is relatively small even at higher ex-posures because of its low charge-collection efficiency,which gives rise to considerable gain-fluctuation noise.Therefore, both high conversion gain and high charge-collection efficiency are required to improve the DQEperformance of an x-ray image detector [45.89]. Theconversion gain depends on W

˙, which is a material

property of the photoconductor. The charge-collectionefficiency can be improved by increasing F and improv-ing the �� products of the carriers. However, increasingF also increases the dark current dramatically in bothHgI2 and CdZnTe detectors. Thus, there is a practicallimitation on F. An F as high as 20V=�m is achiev-able in a-Se detectors while keeping the dark current


E|45.3

within an acceptable level for x-ray imaging [45.38].The charge-collection efficiency of a-Se detectors isrelatively high because of the high F that is used(10�20V=�m) in operating the detector for a reason-ableW

˙.

45.3.3 Modulation Transfer Function (MTF)

Resolution or resolving power is the ability to recordseparate images of small objects that are placed veryclosely together. The overall resolution of a system canbe expressed as a convolution of the component reso-lutions. However, the spatial resolution of an imagingdevice or a system can also be described in terms of theMTF, which is the relative response of the system asa function of spatial frequency. The MTF of an imag-ing system can be described as a cascade of severalstages where the overall MTF is simply the product ofthe MTFs of all the individual stages. The MTF(f ) isa much more convenient descriptor of spatial responsesince the resolving power from multiplication is muchmore easily comprehended than convolution. The over-all MTF (or presampling MTF) of an image detectorcan be expressed as

MTF.f / D MTFm.f /MTFa.f / ; (45.20)

where MTFm.f / is the modulation transfer function ofthe detector material and MTFa.f / is the modulationtransfer function associated with the aperture functionof the pixel electrodes. MTFa.f / arises due to averag-ing of the signal over a pixel area. If the aperture issquare with dimension a, then MTFa.f / will be of theform sinc (af ). The aperture MTF describes how spatialfrequencies are passed through the detector elements.

The spatial resolution in direct-conversion AMFPIis high and closer to the aperture function, as comparedto phosphor-based AMFPI. Based on current under-standing, charge-carrier trapping and reabsorption ofK-fluorescent x-ray photons are the two dominant mecha-nisms responsible for the loss of resolution [45.89] indirect-conversion AMFPI. Some of the K-fluorescentx-ray photons may be reabsorbed at different pointswithin the detector volume from the primary x-ray-photon interaction point. This creates a lateral spreadingof signal and a loss of resolution. The loss of resolutiondue to fluorescence reabsorption is maximum (althoughnot very substantial [45.89]) just above the K-edge ofthe photoconductor. This effect can be ignored when:(i) the incident x-ray photon energy is lower than theK-edge of the photoconductor, or (ii) the mean energyof the x-ray beam and the K-edge occur at widely dif-ferent energies [45.90]. The charge-carrier trapping hasa significant effect on the resolution of these direct-

conversion x-ray image detectors [45.63], which will bediscussed below.

The direct-conversion AMFPI geometry consists ofa photoconductor layer sandwiched between two elec-trodes; the electrode at one side is a continuous metalplate and the electrode on the other side of the pho-toconductor is segmented into an array of individualsquare pixels of size a�a, as shown in Fig. 45.14a. Thegeometric pixel aperture width in a flat-panel detectoris smaller than the pixel pitch (center-to-center spac-ing between two pixels). However, it has been shownthat the effective fill factor (the effective fraction ofpixel area used for image-charge collection) of a photo-conductive flat-panel detector is close to unity [45.91,92]. Therefore, the effective pixel aperture width a isvirtually identical to the pixel pitch. Some of the x-ray-generated carriers are captured by deep traps in the bulkof the photoconductor during their drift across the pho-toconductor layer. Suppose that a carrier is trapped inthe photoconductor above a particular (central) pixelelectrode of a pixelated image sensor. This trappedcarrier induces charges not only on the central pixelelectrode but also on neighboring pixel electrodes, asshown in Fig. 45.14b, and consequently there is a lat-eral spread of information and hence a loss of imageresolution.

An analytical expression for the MTF due to dis-tributed carrier trapping in the bulk of the photocon-ductor has been developed in [45.63], which examinesthe effect of charge-carrier trapping on the resolutionof direct-conversion AMFPI. The effect of trappingon MTF increases with decreasing normalized carrierlifetime (i. e., normalized schubweg). Trapping of thecarriers that move towards the pixel electrodes degradesthe MTF performance, whereas trapping of the othertype of carriers, which move away from the pixels, im-proves the sharpness of the x-ray image.

Figure 45.15 shows the experimental MTF and the-oretical fit of a 300�m-thick positively biased CZTdetector exposed to an 80 kVp x-ray beam with 26mmAl filtration. The experimental data have been extractedfrom [45.93]. The operating F D 0:25V=�m and pixelpitch is 150�m. The Nyquist frequency is 3:3 lp=mm.The dotted line shows the MTF due to the bulk carriertrapping only. The bulk carrier trapping has a signifi-cant effect on the MTF of the detector. As is apparentfrom Fig. 45.15, there is very good agreement be-tween the model and the experimental data. The best-fit�� products of electrons and holes are �e�e D 2:4�10�4 cm2=V and �h�h D 3:2�10�6 cm2=V, which arevery close to the �� values reported previously [45.58,61]. Although the charge-carrier-trapping-limited MTFmodel has been applied to the CZT sensors, the modelcan also be applied to other photoconductive (e.g., a-

PartE|45.3


+–

–

V

F

x L–xx = Lx = 0

a

Charge amplifier

Pixels, bottomelectrodes

Photoconductor

Top electrode

X-rays

x

y

z

Trappedcarrier

a)

b)

Fig. 45.14 (a) A cross section of a direct-conversion pix-elated x-ray image detector. (b) Trapped carriers in thephotoconductor induce charges not only on the centralpixel electrode but also on neighboring pixel electrodes,spreading the information and hence reducing spatial reso-lution

Se, PbO and HgI2) panel x-ray image detectors. Thesame model has been also applied to PbO detectorsin [45.68].

45.3.4 Image Lag and Ghosting

The detectors should be free of image lag and ghostingeffects. When the photoconductor is exposed to x-rays,a transient photocurrent starts to flow that reaches al-most a steady value at the carrier transit time [45.94].Ideally, the current should be present as long as thedetector is subjected to x-ray exposure and, once the ra-diation is stopped, it should diminish to zero. In reality,some of the drifting carriers are trapped in the energydistributed defect states within the bandgap/mobilitygap of the photoconductor layer and these trapped carri-ers are released later. The release time depends on theirenergy depths from the band/mobility edges [45.95]. Asa result, a transient decaying current can be detected forseveral hundreds/thousands of seconds in some detec-tors even after the excitation is removed. This is knownas the residual current or the lag signal, and has beenmodeled for a-Se detectors in [45.94]. For most applica-tions, this is an undesirable phenomenon because whenpart of the current from a previous exposure combines

1.2

1.0

0.8

0.6

0.4

0.2

03.53.02.52.01.51.00.50

MTF

Spatial frequency (lp/mm)

Measured presampling MTFModeled presampling MTFMTF due to trappingPixel aperture function

Fig. 45.15 Measured presampling MTF of a polycrys-talline CdZnTe detector in comparison with modeled re-sults that included blurring due to charge-carrier trappingin the bulk of the photoconductor. The detector thicknessis 300�m and the pixel pitch is 150�m. (Experimentaldata extracted from [45.93, Fig. 12]. Theoretical modelfrom [45.63])

with the next one, the resulting image can be inaccurate.As mentioned in Sect. 45.1.2, the image lag signal wasmeasured in most of the photoconductive detectors. Theamount of image lag signal in a-Se detectors has beenfound to be negligible, whereas it is quite significant insome of other polycrystalline detectors (Sect. 45.1.2).

The x-ray exposure can change the x-ray sensitiv-ity of the exposed area and thus a shadow impressionof a previously acquired image is visible in subsequentuniform exposure, which leads to what is called ghost-ing [45.96]. Ghosting can affect the diagnostic value ofx-ray images in particular when images are acquired ina fast sequence, for example in fluoroscopy. The causesof change in x-ray sensitivity and resolution, and theirrecovery mechanisms in a-Se detectors have been sys-tematically investigated in the last decade [45.97, 98].Though the relative amounts of the sensitivity and res-olution changes owing to repeated exposure in a-Se areunder an acceptable level, they are due to several mech-anisms such as:

1. Recombination of drifting carriers with oppositelycharged trapped carriers

2. Creation of x-ray-induced metastable trap centers,and/or

3. Reduction of the free carrier generation due tospace charge (i. e., due to a nonuniform electricfield) [45.95, 98].

Photoconductors for X-Ray Image Detectors References 1145Part

E|45

It is expected that the ghosting phenomenon mayalso be present in other photoconductive (e.g., HgI2,

CZT, PbO and PbI2) detectors, although it has not yetbeen systematically measured.

45.4 SummaryThe principles of operation of a direct-conversionAMFPI for medical applications have been briefly dis-cussed. The charge-transport and imaging propertiesof some of the potential photoconductors have beencritically discussed and compared with the propertiesof an ideal photoconductor for x-ray image detectors.The various imaging characteristics of photoconductor-based AMFPIs such as the dark current, sensitivity (S),detective quantum efficiency (DQE), and resolution interms of the modulation transfer function (MTF) havealso been examined. These characteristics depend criti-cally not only on the photoconductor’s charge-transportproperties but also on the detector structure, i. e., thesize of the pixel and the thickness of the photocon-

ductor. It has been shown that the detector structure interms of the photoconductor thickness and the pixel sizeis just as important to the overall performance of thedetector as the material properties of the photoconduc-tor itself. Long-term stability, and x-ray-induced effectsand phenomena should also be examined in the futureto obtain more comprehensive detector models.

Acknowledgments. The authors would like to thankDr. John Rowlands for many fruitful discussions andsuggestions during the first edition of this article. Theauthors acknowledge financial support from the NaturalSciences and Engineering Research Council of Canada(NSERC) through its discovery grants program.

References

45.1 J.A. Rowlands, J. Yorkston: Flat panel detectorsfor digital radiography. In: Handbook of Medi-cal Imaging, Vol. 1, ed. by J. Beutel, H.L. Kundel,R.L. Van Metter (SPIE Press, Washington 2000) p. 223

45.2 K.S. Karim, A. Nathan, J.A. Rowlands, S.O. Kasap:IEE Proc.-CDS 150, 267 (2003)

45.3 V.N. Cooper III, T. Oshiro, C.H. Cagnon, L.W. Bassett,T.M. McLeod-Stockmann, N.V. Bezrukiy: Med. Phys.30, 2614 (2003)

45.4 D.C. Hunt, O. Tousignant, J.A. Rowlands: Med. Phys.31, 1166 (2004)

45.5 L.E. Antonuk, J.M. Boudry, Y. El-Mohri, W. Huang,J.H. Siewerdsen, J. Yorkston, R.A. Street: Proc. SPIE2432, 216 (1995)

45.6 J. Chabbal, C. Chaussat, T. Ducourant, L. Fritsch,J. Michaïlos, V. Spinnler, G. Vieux, M. Arques,G. Halm, M. Hoheisel, H. Horbaschek, R. Schulz,M. Spahn: Proc. SPIE 2708, 499 (1996)

45.7 W. Zhao, J.A. Rowlands: Med. Phys. 22, 1595 (1995)45.8 W. Zhao, I. Blevis, S. Germann, J.A. Rowlands,

D. Waechter, Z. Huang: Med. Phys. 24, 1834 (1997)45.9 W. Zhao, J.A. Rowlands: Med. Phys. 24, 1819 (1997)45.10 J.A. Rowlands, S.O. Kasap: Phys. Today 50, 24 (1997)45.11 J.A. Rowlands, W. Zhao, I.M. Blevis, G. Pang,

W.G. Ji, S. Germann, S.O. Kasap, D.F. Waechter,Z.S. Huang: Proc. SPIE 3032, 97 (1997)

45.12 S.O. Kasap, J.A. Rowlands, B. Fogal, M.Z. Kabir,G. Belev, N. Sidhu, B. Polischuk, R.E. Johanson:J. Non-Cryst. Solids 299, 988 (2002)

45.13 W. Zhao, W.G. Ji, A. Debrie, J.A. Rowlands: Med.Phys. 30, 254 (2003)

45.14 D.L. Lee, L.K. Cheung, L.S. Jeromin, E.F. Palecki,B.G. Rodricks: Proc. SPIE 3032, 88 (1997)

45.15 D.L. Lee, L.K. Cheung, B.G. Rodricks, G.F. Powell:Proc. SPIE 3336, 14 (1998)

45.16 O. Tousignant, M. Choquette, Y. Demers, L. Laper-rière, J. Leboeuf, M. Honda, M. Nishiki, A. Taka-hashi, A. Tsukamoto: Proc. SPIE 4682, 503 (2002)

45.17 S.O. Kasap, J.A. Rowlands: J. Mater. Sci. Mater. Elec-tron. 11, 179 (2000)

45.18 M.Z. Kabir, E.V. Emelianova, V.I. Arkhipov, M. Yunus,G. Adriaenssens, S.O. Kasap: J. Appl. Phys. 99,124501 (2006)

45.19 S.O. Kasap, J.A. Rowlands: IEE Proc.-CDS 149, 85(2002)

45.20 S.O. Kasap: Photoreceptors: The chacogenides. In:Handbook of Imaging Materials, 2nd edn., ed. byA.S. Diamond, D.S. Weiss (Marcel Dekker, New York2001) p. 329

45.21 S.O. Kasap, J.A. Rowlands: Proc. IEEE 90, 591 (2002)45.22 F. Greuter, G. Blatter: Semicond. Sci. Technol. 5, 111

(1990)45.23 M.H. Chowdhury, M.Z. Kabir: J. Phys. D 44, 015102

(2011)45.24 A. Zuck, M. Schieber, O. Khakhan, Z. Burshtein: IEEE

Trans. Nucl. Sci. 50, 991 (2003)45.25 S. Tokuda, H. Kishihara, S. Adachi, T. Sato: Proc. SPIE

5030, 861 (2003)45.26 R.A. Street, S.E. Ready, K. Van Schuylenbergh, J. Ho,

J.B. Boyec, P. Nylen, K. Shah, L. Melekhov, H. Her-mon: J. Appl. Phys. 91, 3345 (2002)

45.27 S. Tokuda, H. Kishihara, S. Adachi, T. Sato: J. Mater.Sci. Mater. Electron. 15, 1 (2004)

45.28 M. Simon, R.A. Ford, A.R. Franklin, S.P. Grabowski,B. Mensor, G. Much, A. Nascetti, M. Overdick,M.J. Powell, D.U. Wiechert: IEEE Trans. Nucl. Sci. 52,

PartE|45


188 (2005)45.29 S.O. Kasap, K.V. Koughia, B. Fogal, G. Belev, R.E. Jo-

hanson: Semicond 14, 816 (2003)45.30 I.M. Blevis, D.C. Hunt, J.A. Rowlands: J. Appl. Phys.

85, 7958 (1999)45.31 M.F. Stone, W. Zhao, B.V. Jacak, P. O’Conner, B. Yu,

P. Rehak: Med. Phys. 29, 319 (2002)45.32 S.O. Kasap: J. Phys. D 33, 2853 (2000)45.33 W. Que, J.A. Rowlands: Phys. Rev. B 51, 10500 (1995)45.34 E. Fourkal, M. Lachaine, B.G. Fallone: Phys. Rev. B

63, 195204 (2001)45.35 D. Mah, J.A. Rowlands, J.A. Rawlinson: Med. Phys.

25, 444 (1998)45.36 J.B. Frey, G. Belev, O. Tousignant, H. Mani, L. Laper-

riere, S.O. Kasap: J. Appl. Phys. 112, 014502 (2012)45.37 S.A. Mahmood, M.Z. Kabir, O. Tousignant, H. Mani,

J. Greenspan, P. Botka: Appl. Phys. Lett. 92, 223506(2008)

45.38 B. Polischuk, Z. Shukri, A. Legros, H. Rougeot: Proc.SPIE 3336, 494 (1998)

45.39 S.O. Kasap, J.B. Frey, G. Belev, O. Tousignant,H. Mani, L. Laperriere, A. Reznik, J.A. Rowlands:Phys. Status Solidi (b) 246, 1794 (2009)

45.40 M. Choquette, Y. Demers, Z. Shukri, O. Tousignant,K. Aoki, M. Honda, A. Takahashi, A. Tsukamoto:Proc. SPIE 4320, 501 (2001)

45.41 A. Reznik, S.D. Baranovskii, O. Rubel, G. Juska,S.O. Kasap, Y. Ohkawa, K. Tanioka, J.A. Rowlands:J. Appl. Phys. 102, 053711 (2007)

45.42 K. Kikuchi, Y. Ohkawa, K. Miyakawa, T. Matsubara,K. Tanioka, M. Kubota, N. Egami: Phys. Status Solidi(c) 8, 2800 (2011)

45.43 S.O. Kasap, J.B. Frey, G. Belev, O. Tousignant,H. Mani, J. Greenspan, L. Laperriere, O. Bubon,A. Reznik, G. DeCrescenzo, K.S. Karim, J.A. Row-lands: Sensors 11, 5112 (2011)

45.44 A. Reznik, S.D. Baranovskii, O. Rubel, K. Jandieri,J.A. Rowlands: Phys. Status Solidi (c) 8, 790 (2011)

45.45 S. Abbaszadeh, A. Tari, W.S. Wong, K.S. Karim: IEEETrans. Electron. Devices 61, 3355 (2014)

45.46 S. Abbaszadeh, N. Allec, S. Ghanbarzadeh,U. Shafique, K.S. Karim: IEEE Trans. Electron.Devices 59, 2403 (2012)

45.47 O. Bubon, G. DeCrescenzo, J.A. Raowlands,A. Reznik: J. Non-Cryst. Solids 358, 2431 (2012)

45.48 M.Z. Kabir: J. Mater. Sci. Mater. Electron. 26, 4659(2015)

45.49 G. Juska, K. Arlauskas: Phys. Status Solidi (a) 59, 389(1980)

45.50 N. Hijazi, M.Z. Kabir: Can. J. Phys. 93, 1407 (2015)45.51 Q. Zhao, L.E. Antonuk, Y. El-Mohri, H. Du,

A. Sawant, Z. Su, J. Yamamoto: Med. Phys. 37, 2738(2010)

45.52 H. Jiang, Q. Zhao, L.E. Antonuk, Y. El-Mohri,T. Gupta: Phys. Med. Biol. 58, 703 (2013)

45.53 H. Du, L.E. Antonuk, Y. El-Mohri, Q. Zhao, Z. Su,J. Yamamoto, Y. Wang: Phys. Med. Biol. 53, 1325(2008)

45.54 M. Schieber, A. Zuck, M. Braiman, J. Nissenbaum,R. Turchetta, W. Dulinski, D. Husson, J.L. Riester:IEEE Trans. Nucl. Sci. NS-44, 2571 (1997)

45.55 Z. Su, L.E. Antonuk, Y. El-Mohri, L. Hu, H. Du,A. Sawant, Y. Li, Y. Wang, J. Yamamoto, Q. Zhao:Phys. Med. Biol. 50, 2907 (2005)

45.56 M.Z. Kabir, S.O. Kasap: Appl. Phys. Lett. 80, 1664(2002)

45.57 R.C. Whited, L. Van den Berg: IEEE Trans. Nucl. Sci.NS-24, 165 (1977)

45.58 S.K. Chaudhuri, K. Nguyen, R.O. Pak, L. Matei,V. Buliga, M. Groza, A. Burger, K.C. Mandal: IEEETrans. Nucl. Sci. NS-61, 793 (2014)

45.59 Y. Eisen, A. Shor: Mater. Res. Soc. Symp. Proc. 487,129 (1997)

45.60 T.L. Chu, S.S. Chu, C. Ferekides, C.Q. Wu, J. Britt,C. Wang: J. Appl. Phys. 70, 7608 (1991)

45.61 A. Ruzin, Y. Nemirovsky: J. Appl. Phys. 82, 4166(1997)

45.62 J.G. Mainprize, N.L. Ford, S. Yin, E.E. Gordon,W.J. Hamilton, T.O. Tümer, M.J. Yaffe: Med. Phys.29, 2767 (2002)

45.63 M.Z. Kabir, S.O. Kasap: J. Phys. D 36, 2352 (2003)45.64 G. Zentai, L. Partain, R. Pavlyuchkova, C. Proano,

G. Virshup, L. Melekhov, A. Zuck, B.N. Breen, O. Da-gan, A. Vilensky, M. Schieber, H. Gilboa, P. Bennet,K. Shah, Y. Dmitriev, J. Thomas: Proc. SPIE 5030, 77(2003)

45.65 X. Zhu, H. Sun, D. Yang, J. Yang, X. Li, X. Gao:J. Mater. Sci. Mater. Electron. 25, 3337 (2014)

45.66 R.A. Street, S.E. Ready, F. Lemmi, K.S. Shah, P. Ben-nett, Y. Dmitriyev: J. Appl. Phys. 86, 2660 (1999)

45.67 A. Brauers, U. Schiebel: X-ray image sensor, USPatent (Application) 5729021 (1998)

45.68 M.Z. Kabir: J. Appl. Phys. 104, 074506 (2008)45.69 D.R. Ouimette, S. Nudelman, R. Aikens: Proc. SPIE

3336, 470 (1998)45.70 P.R. Bennett, K.S. Shah, L.J. Cirignano, M.B. Kluger-

man, L.P. Moy, F. Olschner, M.R. Squillante: IEEETrans. Nucl. Sci. 46, 266 (1999)

45.71 S. Yakunin, M. Sytnyk, D. Kriegner, S. Shrestha,M. Richter, G. Matt, H. Azimi, C. Brabec, J. Stangl,M. Kovalenko, W. Heiss: Nat. Photonics 9, 444 (2015)

45.72 T. Baikie, Y. Fang, J.M. Kadro, M. Schreyer, F. Wei,S.G. Mhaisalkar, M. Graetzel, T.J. White: J. Mater.Chem. A 1, 5628 (2013)

45.73 G. Giorgi, J.I. Fujisawa, H. Segawa, K. Yamashita:J. Phys. Chem. Lett. 4, 4213 (2013)

45.74 D.C. Hunt, S.S. Kirby, J.A. Rowlands: Med. Phys. 29,2464 (2002)

45.75 S.A. Mahmood, M.Z. Kabir: Curr. Appl. Phys. 9, 1393(2009)

45.76 Z. He: Nucl. Instr. Meth. A 463, 250 (2001)45.77 R.A. Street: Appl. Phys. Lett. 57, 1334 (1990)45.78 M.Z. Kabir: Nucl. Instr. Meth. A 736, 156 (2014)45.79 S.A. Mahmood, M.Z. Kabir: J. Vac. Sci. Tech. A 29,

031603 (2011)45.80 G. Vincent, A. Chantre, D. Bois: J. Appl. Phys. 50,

5484 (1978)45.81 X. Zhu, H. Sun, D. Yang, J. Yang, X. Li, X. Gao:

J. Mater Sci. Mater. Electron. 25, 3337 (2014)45.82 M.Z. Kabir, S.O. Kasap: J. Vac. Sci. Tech. A 20, 1082

(2002)45.83 E. Rosencher, V. Mosser, G. Vincent: Phys. Rev. B 29,

1135 (1984)

Photoconductors for X-Ray Image Detectors References 1147Part

E|45

45.84 J.M. Boone: X-ray production, interaction, and de-tection in diagnostic imaging. In: Handbook ofMedical Imaging, Vol. 1, ed. by J. Beutel, H.L. Kun-del, R.L. Van Metter (SPIE Press, Washington 2000)p. 1

45.85 M.Z. Kabir, S.O. Kasap: J. Vac. Sci. Tech. A 22, 975(2004)

45.86 M.Z. Kabir, S.O. Kasap: J. Phys. D 35, 2735 (2002)45.87 M. Rabbani, R. Shaw, R.L. Van Metter: J. Opt. Soc.

Am. A 4, 895 (1987)45.88 J.G. Mainprize, D.C. Hunt, M.J. Yaffe: Med. Phys. 29,

976 (2002)45.89 M.Z. Kabir, S.O. Kasap, W. Zhao, J.A. Rowlands: IEE

Proc.-CDS 150, 258 (2003)45.90 W. Que, J.A. Rowlands: Med. Phys. 22, 365 (1995)45.91 D.C. Hunt, O. Tousignant, Y. Demers, L. Laperrière,

J.A. Rowlands: Proc. SPIE 5030, 226 (2003)

45.92 G. Pang, W. Zhao, J.A. Rowlands: Med. Phys. 25,1636 (1998)

45.93 S. Tokuda, H. Kishihara, S. Adachi, T. Sato: J. Mater.Sci. Mater. Electron. 15, 1 (2003)

45.94 S. Siddiquee, M.Z. Kabir: J. Vac. Sci. Tech. A 33,041514 (2015)

45.95 S.A. Mahmood, M.Z. Kabir, O. Tousignant,J. Greenspan: IEEE Trans. Nucl. Sci. 59, 597(2012)

45.96 V. Loustauneau, M. Bissonnette, S. Cadieux,M. Hansroul, E. Masson, S. Savard, B. Polischuk:Proc. SPIE 5368, 162 (2004)

45.97 M.Z. Kabir, M. Yunus, S.O. Kasap: Proc. SPIE 5368,170 (2004)

45.98 M.Z. Kabir, L. Chowdhury, G. DeCrescenzo, O. Tou-signant, S.O. Kasap, J.A. Rowlands: Med. Phys. 37,1339 (2010)

Documents

45.PhotoconductorsforX-RayImageDetectors Photoconduc · 2017-10-04 · Multiplexer Scan timing Scanning control Fig.45.2 Schematic diagram that shows some pixels of the active matrix