Lecture Notes 2

Charge-Coupled Devices (CCDs) – Part I

• Basic CCD Operation

• CCD Image Sensor Architectures
• Static and Dynamic Analysis
◦ Charge Well Capacity
◦ Buried channel CCD
◦ Transfer Efficiency
◦ Readout Speed

EE 392B: CCDs – Part I 2-1

 Preliminaries

• Two basic types of image sensors: CCD and CMOS

• Photodetector elements are similar
◦ Photodiode
◦ Photogate
◦ Pinned-diode
The photodiode and photogate operation were covered in Lecture Notes
1. We discuss the pinned diode in Part II of this lecture notes
• Main difference is the readout “circuitry” / mechanism
◦ In CCDs, charge is shifted out
◦ In CMOS image sensors, charge or voltage is read out using row and
column decoders — similar to a random access memory

EE 392B: CCDs – Part I 2-2

 • The readout circuits (including in pixel devices) determine the sensor
conversion gain, which is the output voltage per electron collected by the
photodetector, in µV/electron
• Given the sensor quantum efficiency, conversion gain, and area its
sensitivity measured in V/Lux·s can be determined
• Readout speed determines the video frame rate that an image sensor can
operate at – 30 to 60 frames/s are typical, but lower frame rates are
sometimes dictated by the available bandwidth (e.g., wireless camera),
and higher frame rates are required for many industrial and military
applications

EE 392B: CCDs – Part I 2-3

 Charge-Coupled Devices

• CCD is a dynamic analog (charge) shift register

• It consists of a series of MOS capacitors coupled with one another
• CCD is clocked, and all operations are in transient mode
• Charge is coupled from one gate to the next gate by fringing electric field,
potential and carrier density gradient
• We first discuss CCD operation using different clocking methods, then
discuss their use in image sensors

EE 392B: CCDs – Part I 2-4

 Potential Well Analogy – MOS Capacitor

EE 392B: CCDs – Part I 2-5

 Basic 4-Phase CCD

• A uni-directional potential gradient must be provided for charge transfer

• Gap between electrodes must be small (compared to the oxide thickness)
to enable fringing electric field from neighboring electrodes to couple the
charge across

EE 392B: CCDs – Part I 2-6

 Surface Potential Control

EE 392B: CCDs – Part I 2-7

 CCD Electrodes

• Overlapping polysilicon electrodes

◦ Oxidized polySi sidewall provides the electrode isolation
◦ 2 levels polySi for 4-phase and 2-phase CCD
◦ 3 levels polySi for 3-phase CCD
• Single polySi electrode also possible provided the electrode gap is small
• Most common: 4-phase, 3-phase, 2-phase
◦ Other uncommon ones: ripple clock, accordion clock

EE 392B: CCDs – Part I 2-8

 Charge Transfer in a 4-Phase CCD

A. Theuwissen, “Solid State Imaging with Charge-Coupled Devices,” Kluwer (1995)

EE 392B: CCDs – Part I 2-9

 Charge Transfer in a 3-Phase CCD

A. Theuwissen, “Solid State Imaging with Charge-Coupled Devices,” Kluwer (1995)

EE 392B: CCDs – Part I 2-10

 Charge Transfer in a 2-Phase CCD

A. Theuwissen, “Solid State Imaging with Charge-Coupled Devices,” Kluwer (1995)

EE 392B: CCDs – Part I 2-11

 Frame Transfer CCD Image Sensor

Integration
Light−sensitive
CCD array Vertical shift

Vertical shift

Horizotal shift
Frame−store
Time
CCD array

Amplifier Operation
Horizontal CCD Output

• Top CCD array used for photodetection (photogate) and vertical shifting
• Bottom CCD array optically shielded – used as frame store
• Operation is pipelined: data is shifted out via the bottom CCDs and the
horizontal CCD during integration time of next frame
• Transfer from top to bottom CCD arrays must be done very quickly to
minimize corruption by light, or in the dark (using a mechanical shutter)
• Output amplifier converts charge into voltage, determines sensor
conversion gain
EE 392B: CCDs – Part I 2-12
 Interline Transfer CCD Image Sensor

Vertical CCD
Integration

Transfer

Vertical shift

Horizotal shift
Photodiode
Time

Operation

Horizontal CCD Output

Amplifier

• Photodiodes or pinned diodes are used

• All CCDs are optically shielded, used only for readout
• Collected charge is simultaneously transferred to the vertical CCDs at the
end of integration time (a new integration period can begin right after the
transfer) and then shifted out
• Charge transfer to vertical CCDs simultaneously resets the photodiodes,
(shuttering done electronically for “snap shot” operation)
EE 392B: CCDs – Part I 2-13
 Frame Transfer Versus Interline Transfer

• Frame transfer uses simpler technology (no photodiodes), and achieves

higher fill factor, which is the fraction of pixel area occupied by the
photodetector, than interline transfer
• Interline transfer uses optimized photodiodes with better quantum
efficiency than the photogates used in frame transfer
• In interline transfer the image is captured at the same time (“snap shot”
operation) and the charge transfer is not subject to corruption by
photodetection (can be avoided in frame transfer using a mechanical
shutter)
• Frame transfer has shorter integration time for the same frame rate than
interline due to its nonoverlapping integration and readout times
• Frame transfer chip area (for the same number of pixels) can be larger
than interline transfer
• Most of today’s CCD image sensors use interline transfer

EE 392B: CCDs – Part I 2-14

 Surface Channel CCD Static Analysis

• Let’s examine the amount of charge a surface channel CCD can store
• We need to relate the CCD gate voltage to the surface potential when
there is depletion charge and mobile charge (Qs in electrons/cm2) under
the gate (MOS capacitor)
• Surface potential under an MOS capacitor:

EE 392B: CCDs – Part I 2-15

 • We can relate the surface potential ψs to the applied bias voltage vG and
the charge stored Qsig in electrons/cm2 as follows
The gate voltage is given by
qQs
vG = ψ s − v F B + ,
Cox
where
Qs = Naxd + Qsig
is the total charge
The depletion width is given by (see Appendix III of LN 1)
s
2sψs
xd = ,
qNa
where q
ψs = v1 + v2 − v22 + 2v1v2
Here
qQsig
v1 = v G + v F B +
Cox
2
v2 = sqNa/Cox

EE 392B: CCDs – Part I 2-16

 Well Charge Capacity for Surface Channel CCD

• Empty well – semiconductor is depleted

• Non-empty well – semiconductor is less depleted or has mobile charge

EE 392B: CCDs – Part I 2-17

 • We can derive the well charge capacity Qwell in electrons/cm2 using the
equations above as follows.
First at the top of the potential wells, ψs1 = ψs2, thus
sqNa
vG1 + vF B1 = ψs1 +
Cox
sqNa qQwell
vG2 + vF B2 = ψs1 + + ,
Cox Cox
Now, assuming vF B1 = vF B2 and subtracting, we obtain
qQwell
vG2 − vG1 =
Cox
Thus
(vG2 − vG1)Cox
Qwell =
q
The excess gate voltage is balanced by the charge in the well

EE 392B: CCDs – Part I 2-18

 Surface Channel CCD Static Design Curves

A. Theuwissen, “Solid State Imaging with Charge-Coupled Devices,” Kluwer (1995).

EE 392B: CCDs – Part I 2-19

 Buried Channel CCD

• Having the mobile charge at the surface as in the surface channel CCD
(SCCD) has several disadvantages
◦ Si/SiO2 interface states will capture the mobile carriers and release
them at a later time, causing transfer inefficiency (to be discussed
later)
◦ Si/SiO2 interface states introduce dark current due to surface
generation (as discussed in Lecture Notes 1)
• It is therefore desirable to store the carriers some distance below the
Si/SiO2 interface – this is called buried channel CCD (BCCD)
• In a BCCD, the surface is doped with the opposite polarity (e.g., n-Si if
the substrate is p-Si), forming a junction beneath the surface
• The surface doping region (n-Si) is completely depleted (empty well)
• As signal charge is stored in the BCCD, the surface doping region
becomes less depleted

EE 392B: CCDs – Part I 2-20

 Channel Potential of SCCD and BCCD

EE 392B: CCDs – Part I 2-21

 Buried Channel CCD Static Characteristics

• In order to determine the charge well capacity of a BCCD, we need to

know the minimum of the energy well (or maximum of potential well)
φmin, where φmin is referenced to the substrate potential
• To avoid surface channel operation, the energy minimum should be a few
kT below the potential at the surface
• We calculate the relationship between the applied voltage vG, the signal
charge Qsig , and the potential φmin in a way analogous to the surface
channel case
• By KVL,
vG + vF B + vox + φs = φmin and φmin = φj + φc
By Gauss’s law
q(Nd(t − ∆t) − Qsig )
vox =
Cox

EE 392B: CCDs – Part I 2-22

 Using standard pn junction theory, we can relate the potential across the
depletion region to its thickness as follows
 2
qNd Qsig
φs = t − ∆t − ,
2s Nd
qNax2p qNd(∆t)2
φj = , φc = , and
2s 2s
Nd∆t
xp =
Na
Note that φs is the same as ψs. Combining the above equations, we
1/2
obtain the following quadratic equation in φmin
   12
Nd 1 t Qsig 2qsNaNd 1
2
φmin + + − φmin = vG + vF B
Nd + N a Cox s Nds  Nd + Na 
1 t 1 t Qsig
+qNdt + − qQsig + −
Cox 2s Cox s 2Nds
After finding φmin, all other unknowns (φs, φj , φc, ∆t, vox, and xp) can be
found
Reference: D. F. Barbe, “Imaging Devices Using the Charge-Coupled Concept,” Proceedings of IEEE, vol 63,
pp 28-67, 1975

EE 392B: CCDs – Part I 2-23

 Buried Channel CCD Static Design Curves

A. Theuwissen, “Solid State Imaging with Charge-Coupled Devices,” Kluwer (1995).

EE 392B: CCDs – Part I 2-24

 Charge Transfer in CCD

• Four mechanisms are involved in the dynamic charge transfer process

◦ Self-induced drift
◦ Electrostatic repulsion
◦ Thermal diffusion
◦ Fringing field
• The first charge carriers are transferred by self induced drift and the
charge is re-distributed within the well by electrostatic repulsion Both
processes are fairly quick.
• Thermal diffusion is a slower process.
• Finally, the last remaining charge is transferred by fringing field
• The charge transfer efficiency η is the fraction of charge transferred (1 − η
is the inefficiency)
• Self-induced drift and electrostatic repulsion give 1 − η ≈ 10−2

EE 392B: CCDs – Part I 2-25

 • Adding thermal diffusion gives 1 − η ≈ 10−3
• Further adding fringing field gives 1 − η ≈ 10−4
• In order to achieve high transfer efficiency and high transfer speed, CCD
should be designed to have large fringing field

EE 392B: CCDs – Part I 2-26

 Charge Transfer in CCD

• The four driving forces involved in the charge transfer process can be
quantified by an equivalent electric field. Define the signal charge areal
density as ρs(y, t) in electrons/cm2
◦ Self-induced drift:
q ∂ρs
Ed = ·
Cox ∂y
◦ Electrostatic repulsion:
2qtox ∂ρs
Er = − · ,
ox + s ∂y
◦ Thermal diffusion:
kT ∂ρs
Eth = ·
qρs ∂y
◦ Fringe field: EF R
• We provide a qualitative explanation of these four quantities

EE 392B: CCDs – Part I 2-27

 Self-Induced Drift

• A gradient in charge distribution (during the transfer process) results in a

gradient in surface potential along the transfer direction (y)

• Assuming tox << xd, we obtain

q ∂ρs
Ed ≈ ·
Cox ∂y

EE 392B: CCDs – Part I 2-28

 Electrostatic Repulsion

• Electrostatic repulsion arises from the electrostatic forces acting on q(y)

by q(y 0) (repulsion) and its image charge (attraction)

• The resulting field is given by

2qtox ∂ρs q ∂ρs
Er = − · ≈− · ,
ox + s ∂y 2Cox ∂y
where we made the approximation s ≈ 3ox
EE 392B: CCDs – Part I 2-29
 Thermal Diffusion

• A gradient of charge density along the transfer direction (y) establishes a

diffusion current
• The diffusion can be modeled as an effective electric field acting on the
carriers, which is given by
kT ∂ρs
Eth = ·
qρs ∂y
• This effective field depends on carrier density in the CCD well

EE 392B: CCDs – Part I 2-30

 Fringe Electric Field

• Fringe electric field arises from the 2-dimensional nature of the potential
in the device – coupling from neighboring electrodes
• The fringe electric field is smallest at the middle of the electrode

EE 392B: CCDs – Part I 2-31

 • The following approximate analytical expression for EF R(min) can be found
by solving the 2D Poisson equation
 4
2π s∆v 5xdL
EF R(min) ≈
3 L2Cox 1 + 5xdL
where
∆v = vG2 − vG1
J. E. Carnes, W. F. Kosonocky, and E. G. Ramberg, ”Drift-aiding fringing fields in charge-coupled devices,” IEEE
Journal of Solid-State Circuits, vol. 6, pp. 322 - 326, October 1971.

EE 392B: CCDs – Part I 2-32

A. Theuwissen, “Solid State Imaging with Charge-Coupled Devices,” Kluwer (1995).

EE 392B: CCDs – Part I 2-33

 Methods to Increase Fringing Field

• Increase gate oxide thickness

• Decrease substrate doping
• Reduce CCD gate length (EFR varies as 1/L2)
• Increase the gate voltage difference

EE 392B: CCDs – Part I 2-34

 Charge Transfer Advantage of BCCD

• A major advantage of buried channel CCD is the large fringing field,

which results in fast charge transfer

EE 392B: CCDs – Part I 2-35

 Charge Transfer Efficiency

• The CCD charge transfer efficiency, η ≤ 1, is the fraction of signal charge

transferred from one CCD stage to the next, i.e.,
Q(t)
η =1− ,
Q(0)
where Z L
Q(t) = W ρs(y, t) dy
0
Here L and W are the gate length and width of the CCD, respectively
• η must be made very close to 1, because in a CCD image sensor charge is
transferred up to n + m CCD stages (3 × (n + m) times for 3-phase CCD)

EE 392B: CCDs – Part I 2-36

 • Example: consider a 1024 × 1024 CCD image sensor
The table lists the charge transfer efficiency η and the corresponding
worst case fraction of charge transferred to the output

η fraction at output
0.999 0.1289
0.9999 0.8148
0.99999 0.9797

EE 392B: CCDs – Part I 2-37

 Transfer Efficiency Analysis

• The “equivalent” electric field that acts on the charge carriers in a CCD
well can be used to derive the charge transfer efficiency
• The derivations are too detailed to go through in this course, so we just
state the results
• The total equivalent electric field is given by
Etot(y, t) = Ed + Er + Eth + EF R
q ∂ρs q ∂ρs kT ∂ρs
≈ · − · + · + EF R
Cox ∂y 2Cox ∂y qρs ∂y
q ∂ρs kT ∂ρs
≈ · + · + EF R
2Cox ∂y qρs ∂y
• Recall that the current density can be expressed as
∂n
j = qµnnE + qD
∂y

EE 392B: CCDs – Part I 2-38

 We can then transform the equivalent electric field into an equivalent
diffusion constant as if the current is all due to diffusion
 
 ρs kT EF R(y) 
Def f = µn  + +  
2Cox q 1 ∂ρs
·
ρs ∂y

Now, we weight average over the length of the CCD gate to obtain
2LE
 
ρs kT F R(min)
Def f = µn + +
2Cox q π
• The transfer inefficiency is thus ∝ e−t/τ , where
1  π 2
= Def f ×
τ 2L
• High transfer efficiency is achieved if
◦ the time available to complete the transfer is long enough (a
function of CCD clocking speed)
◦ surface state density is low
• Since τ ∝ L2, shorter CCD gates will have faster transfer

EE 392B: CCDs – Part I 2-39

 A Simple Illustration

• 99% of charge transferred immediately and remaining 1% transferred

slowly by thermal diffusion and fringing field
• Last 1% accounts for most of the transfer time

A. Theuwissen, “Solid State Imaging with Charge-Coupled Devices,” Kluwer (1995).

EE 392B: CCDs – Part I 2-40

 CCD Readout Speed

• CCD imager readout speed is limited mainly by the array size and the
charge transfer efficiency requirement
• Example: consider a 1024 × 1024 3-phase interline transfer CCD image
sensor with η = 0.99997, L = 4µm, and Dn = 35cm2/s, find the maximum
video frame rate
transfer time for the horizontal CCD limits the readout speed
to find the minimum required transfer time per CCD stage, tmin, we use
the equation
t
− min
η = 0.99997 = (1 − 0.01e 3τ )3,
which gives tmin = 37.8ns, thus the time required to shift one row out is
37.8ns×1024 = 38.7µs
ignoring the row transfer time (from vertical CCDs), we get
minimum frame transfer time of 39.6ms, or maximum video frame rate of
25frame/s
Note: CMOS image sensors can be much faster than CCDs

EE 392B: CCDs – Part I 2-41