UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede LECTURE NOTES 5 Materials made up of normal matter (atoms, molecules, etc.) have some amazing electromagnetic properties! Simplest kinds of electromagnetic properties: A.) conductor (of electricity) ⎫ ⎪ B.) partial conductor/insulator ⎬ C.) non-conductor ⇒ insulator ⎭⎪ Why materials conduct vs. do not conduct electricity depends on microscopic (i.e. quantum) structure of materials & temperature (i.e. thermal/internal energy). CONDUCTORS: "normal" good conductors of electricity: ⎞ ⎟ metals - gold, platinum, silver, copper... ⎠ Have finite DC resistance, R = V/I (Ohm’s Law) @ finite temperatures, T > 0 K. "superconductors" - low temperature SC's such ⎞ ⎟ as lead (Tc ~4K) indium, niobium, ..... ⎟ Hi-Tc SC's (e.g. Tc ~77K): BSCO, YBCO ..... ⎟⎠ DC resistance vanishes below Tc (critical temp) INSULATORS: e.g. plastics, teflon, glass, rubber …. PARTIAL CONDUCTORS: e.g. doped plastics, semi-conductors (germanium, silicon, graphite….) IONIC LIQUIDS: e.g. salt water – can also conduct electricity Acidic solutions – ions transport electrical charges – not electrons An ideal/perfect conductor is a (hypothetical) material that would have an unlimited number of completely free electrons/free charges. No such things truly exist in nature, but ∃ many materials which do come (amazingly) close to an ideal/perfect conductor. ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. 1 UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 One important property of a conductor is that: Prof. Steven Errede n.b. ENET ( r ) ≡ 0 is a different physics statement than E ( r ) ≡ 0 inside a conductor! ENET ( r ) ≡ 0 inside a conductor 1) Lect. Notes 5 If ENET ( r ) = 0 inside a conductor, then free charges inside the conductor would move/be accelerated by this Einside ( r ) = 0 , because: Finside ( r ) = qe Einside ( r ) = me ae ( r ) . Indeed, this is precisely what happens inside a conductor e.g. when it is placed in a uniform external electric field Eext ( r ) = E0 xˆ – the free charges inside the conductor (electrically neutral!) re-distribute themselves to create/produce Einside ( r ) = 0 on extremely short timescales of ~ femto → pico-seconds ( ~ 10−15 − 10−12 sec ) . The redistributed free charges pile up on the surface(s) of the conductor in such a way as to produce Einside ( r ) = 0 . These induced charges produce an internal electric field of their own, which exactly cancels the external field, Eext ( r ) ! e.g. consider a block of metal in a uniform external electric field, Eext ( r ) = E0 xˆ Enet inside ( r ) = Eext ( r ) + Einduced ( r ) = 0 inside ⇒ Einduced ( r ) = − Eext ( r ) = − E0 xˆ inside Note that: Enet outside ( r ) = Eext ( r ) when a conductor is placed inside an initially uniform applied field, Eext ( r ) Induced surface charge, σ Eext ( r ) = E0 xˆ 2 Induced surface charge, σ ⊕ ⎧ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎨ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎩ E0 xˆ ← ⊕ ← ⊕ ← ⊕ ← ⊕ ← ⊕ E (r ) ⊕ inside ⊕ ← ⊕ ← ⊕ ← ⊕ ← ⊕ ⎫ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎬ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎪ ⎭ x̂ Eext ( r ) = E0 xˆ ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede Another important property of a conductor is that: free 2) The volume free charge density, ρinside ( r ) = 0 inside a conductor. This follows from Gauss’ Law (differential form): free ∇ ⋅ Einside ( r ) = ρinside (r ) εo but if: Einside ( r ) = Eext ( r ) + Einduced ( r ) = 0, ∀r (i.e. everywhere inside conductor) inside then: ∇ ⋅ Einside ( r ) = 0, ∀rinside free ⇒ ρinside ( r ) = 0, ∀rinside 3) Any induced charges on a conductor can ONLY reside on surface(s) of the conductor – as surface charge distributions, σ free (n.b. free surface charges minimize overall potential energy (i.e. maximize overall entropy of system!)) free If ∃ induced free charges, and ρinside ( r ) = 0 ∀r , the only place(s) such induced free charges can reside is on the surface(s) of the conductor, as σ free ( r ) . 4) The entire volume & surface of a conductor is an equipotential. If a & b are two arbitrary points, ra & rb on the surface of a conductor, the must!! potential difference, ΔVab ≡ Vb − Va = V ( rb ) − V ( ra ) = − ∫ E ( r )idl = 0 ⇒ Va = Vb b a If ΔVab ≠ 0 then free charges will move!! ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. 3 UIUC Physics 435 EM Fields & Sources I 5) Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede Just outside the surface of a conductor, Eoutside ( r @ surface ) is perpendicular/normal to the surface, i.e. Eoutside ( r @ surface ) nˆsurface If ∃ an E ( r ) = Etangential ( r ) @ surface, ⇒ free charge will move/flow!!! This can’t happen in electrostatics!! ⎛ Coulombs ⎞ Example: A conducting surface of total surface area, Acond is charged with σ free = σ o ⎜ ⎟. m2 ⎝ ⎠ What is the electric field strength (i.e. electric field intensity) at the surface of the conductor, Esurface ? ⇒ Use (the integral form of) Gauss’ Law – choose fully enclosing Gaussian surface S to be infinitesimally above the conducting surface (e.g. δ h larger than conducting surface everywhere). Then take limit as δ h → 0 : σ free Acond Q Esurface ( r ) nˆ everywhere. ∫ E idA = encl = S ε0 = Esurface Acond ε0 ∴ Esurface = Esurface = σ free σ o = (Volts m ) εo εo n.b. If we had instead used e.g. a “shrunken” Gaussian pillbox on surface of conductor: Q =0 =0 =0 ∫ E idA = ∫ Esurface idA1 + ∫ E tangential idA2 + ∫ Einside idA3 = encl S S1 π R 2 Esurface = S2 Qencl ε0 S3 = σ oπ R 2 ⇒ εo Esurface = εo σ free σ nˆ = o nˆ εo εo Get same answer! (We should!) 4 ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede The Free Surface Charge σ free The Surface Electric field, E surface ( r ) The Surface Potential, V ( r ) , Electrostatic Force & Pressure (Force per Unit Area) Acting on a Conductor We have derived, using Gauss’ Law: Esurface ( r ) = σ free = Esurface ( r ) εo In vector notation: Esurface ( r ) = σ free nˆ εo Where n̂ is the outward pointing unit normal vector (outward = by convention). From Griffiths Eqn’s 2.34-2.37, p. 89-90: n̂ Eabove ⎡⎣Vbabove ( rb ) − Vabelow ( ra ) ⎤⎦ lim δ h→0 b b a a = − ∫ E i d = ∫ ∇V i d = 0 ( surface of conductor is an equipotential ) b ½ δh ½ δh Gave: ∇V above ( r ) − ∇V below ( r ) a Ebelow surface of conductor OR: =0 =− σ free nˆ ε0 σ ∂V ( r ) ≡ ∇V ( r )inˆ = − free surface ∂n surface εo OR: σ free = − ε o ∂V ( r ) ∂n surface If Esurface ( r ) or Vsurface ( r ) is known, then σ free can be obtained (and/or vice-versa). ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. 5 UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede FORCE & PRESSURE ON A CONDUCTOR ( Consider an arbitrarily-shaped conductor, with σ free coulombs ) free charge residing on its m2 surface. Infinitesimally small surface “patch,” of surface area Apatch = π R 2 . n̂ • R Apatch = π R 2 Now Apatch also has surface charge σ free on it. The total charge of this patch is Qpatch = π R 2σ free . Edge-on view of infinitesimally small patch: above nˆ , E patch = σ free σ ( + nˆ ) = + free nˆ 2ε o 2ε o σ free on Apatch ( = π R 2 ) +++++++ E below patch = σ free σ ( −nˆ ) = − free nˆ 2ε o 2ε o We have discussed before (Griffiths Ex. 2.4, p. 73-74 and/or P435 Lect. Notes 2 p. 9-12) that a surface charge σ has a (net) E -field ⊥ to surface on which σ resides, both above and below the surface. The transverse/tangential E -field components (on a flat surface, from symmetry) were shown to cancel. (n.b. an infinitesimally small surface patch is FLAT.) But we also said that Einside ( r ) = 0 inside a conductor! How do we reconcile these two statements? Consider an edge-on view of the (arbitrarily shaped conductor, but with the (infinitesimally) small surface patch (of area Apatch) removed: σ free on surface of conductor 6 ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede Now consider just the electrostatic surface charge itself – i.e. mentally “erase” the presence of Enet ( r ) circular “hole” where infinitesimal the conductor altogether: n̂ patch of area Apatch = π R 2 was (now removed) The surface charge distribution is a (very!) special configuration that is an equipotential surface – but it is still a collection of individual point charges (@ the microscopic level). Each individual point charge q has its own radial electric field, Epoint ( r ) E point q We ask: what is the net electric field Enet ( r ) on the conductor’s surface? (e.g. at the location of the “hole” where the patch of infinitesimal area Apatch = π R 2 was (which is now removed)). It’s hard to rigorously prove the following, for an arbitrarily-shaped charged conductor, but it can be rigorously proved (analytically) for symmetrically-shaped conductors – e.g. a charged sphere, with an infinitesimally small spherical cap removed, e.g. at its north pole. (see e.g. J.D. Jackson’s book Electrodynamics for this proof…) In general, for arbitrarily-shaped conductors, because their charged surfaces are equipotential surfaces, the net electric field, Enet ( rpatch ) on the conductor’s surface, right at the “hole” location of the (missing/removed) patch, associated with the free charge everywhere else on the conductor’s surface, is: Enet ( rpatch ) = + σ free nˆ 2ε o (!!!) infinitesimally above/below surface, @ the patch “hole” location: above Enet ( rpatch ) = Enet ( rpatch ) = Enetbelow ( rpatch ) = + Thus, we see that: above outside above EToT ( rpatch ) = E patch ( rpatch ) + Enetabove ( rpatch ) above EToT ( rp ) = And we also see that: σ free nˆ 2ε 0 σ free σ σ nˆ + free nˆ = free nˆ 2ε o 2ε o εo below inside below EToT ( rpatch ) = E patch ( rpatch ) + Enetbelow ( rpatch ) below EToT ( rpatch ) = − σ free σ nˆ + free nˆ = 0 2ε o 2ε o ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. 7 UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede Thus, the net electric field just above the surface of a conductor arises from two equal contributions – the E patch from σ free on the infinitesimal patch of area Apatch = π R 2 itself, and the net contribution from the free charge on the remainder of the conducting surface! Likewise, the internal E field inside the conductor = 0, because these two contributions cancel each other!!! Thus, VECTORIALLY, in the region of the infinitesimally small patch of area Apatch = π R on/at the surface of the conductor, we have: n̂ above outside EToT ( rpatch ) = σ free nˆ εo Apatch • below inside EToT (r ) = 0 above outside above EToT ( rpatch ) = E patch ( rpatch ) + Enetabove ( rpatch ) above EToT ( rp ) = below inside below EToT ( rpatch ) = E patch ( rpatch ) + Enetbelow ( rpatch ) σ free σ σ nˆ + free nˆ = free nˆ εo 2ε o 2ε o below EToT ( rpatch ) = − σ free σ nˆ + free nˆ = 0 2ε o 2ε o total We are now in a position to ask: what is the net/total force, Fpatch ( r ) acting on the infinitesimally small patch, of area Apatch = π R 2 ? total 1) There can be no contribution(s) to the net/total force, Fpatch ( r ) acting on the patch due to the above free surface charge, σ free (or associated E − fields E patch ( r ) + E below patch ( r ) ) acting on the patch itself (analogous to trying to lift yourself up by pulling on your shoes!) total Fpatch ( rpatch ) = Qpatch * Eat patch ( rpatch ) = Qpatch * Enet ( rpatch ) net E -field @ patch from σ free elsewhere on conducting surface ⎛σ total Fpatch ( rpatch ) = σ free Apatch * Enet ( rpatch ) = σ free Apatch * ⎜ 2εfree ⎝ o 8 ⎞ nˆ ⎟ ⎠ ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. UIUC Physics 435 EM Fields & Sources I total patch F (r ) = patch σ 2free ( rpatch ) 2ε o Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede NOTE: this is a force which points in the outward normal (⊥) direction! Apatch nˆ If we sum up the infinitesimal force contributions from all the “patches” associated with the entirety of the conducting surface, the magnitude of the force is: N σ 2free ( rpatch ) σ 2free n.b. Assumption made here total Fconductor ( rsurface ) = ∑ Fpatch ( rpatch ) = ∫ dA = Aconductor conductor is that σ free ( r ) = constant 2ε o 2ε o i =1 Now, pressure ≡ force/unit area, i.e. P = F Thus: Since: A 2 1 ⎛ σ free ( rsurface ) ⎞ ε o 2 ⎜ ⎟ = E@ surface ( rsurface ) Then, electrostatic pressure, P@ surface ( rsurface ) = ε o 2 ⎟ 2 2 ⎜ ε o ⎝ ⎠ Esurface ( rpatch ) = 2 Enet ( rpatch ) = 2 σ free ( rpatch ) 2ε o ⎛ σ free ( rpatch ) ⎞ ⎟ nˆ nˆ = ⎜ ⎜ ⎟ ε o ⎝ ⎠ 2 1 ⎛ σ free ( rsurface ) ⎞ ε o 2 ⎟ = E@ surface ( rsurface ) Then, electrostatic pressure, P@ surface ( rsurface ) = ε o ⎜ 2 ⎟ 2 2 ⎜ ε o ⎝ ⎠ n.b. If electrostatic forces are able/capable of performing mechanical work W, then must do so at the expense of electrostatic energy (recall that total energy must always be conserved!) n.b. metals are elastic solids!! ⇒ Suppose a small area, a of the conductor is pulled into the electric field region by an infinitesimal distance, Δx . Then work done by field on area, a is: F area, a Δx ⎛ F (r ) ⎞ ΔW ( r ) = F ( r )iΔx = ⎜ ⎟ ( aΔx ) = P@ surface ( r ) * aΔx = P@ surface ( r ) ΔV a ⎠ = volume, ΔV ⎝ F Δx OR: ∴ ΔW ( rsurface ) = P@ surface ( rsurface ) ΔV P@ surface ( rsurface ) = ΔW ( rsurface ) ΔV = Energy Density, uE ( rsurface ) Electrostatic Pressure, P (@ surface) = Volume Energy Density of Electrostatic Field, uE (@surface) 1 Joules 3 P@ surface ( r @ surface ) = U E @ surface ( r @ surface ) = ε o E@2 surface ( r @ surface ) m 2 n.b. Electrostatic field exerts a negative pressure on conductor – i.e. electrostatic force on the conductor pulls conductor into electrostatic field !!! ( ) ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. 9 UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede CAPACITORS A capacitor is a device that enables the storage (long and/or short term) of electric charge, Q. Since there are electric fields associated with electric charge, a capacitor is also a device that enables the storage (long and/or short term) of electrical energy. One can make a simple capacitor using e.g. two parallel, very thin, conducting plates, separated by a distance d and initially uncharged. Then, we put charge +Q e.g. on the upper plate and put charge –Q on the lower plate. A potential difference, ΔVab = Vb ( z = d ) − Va ( z = 0 ) equipotential surface now exists between the two plates, as shown in the figure below: Vb ( z = d ) point b +Q on upper plate +++++++++++++++++ + + +++++++++++++++++ E (r ) Potential Difference ΔVab = Vb ( z = d ) − Va ( z = 0 ) z=d d = plate separation −−−−−−−−−−−−−−−−−−−− Equipotential Surface Va ( z = 0 ) − −−−−−−−−−−−−−−−−−−−− − point a z=0 −Q on lower plate A static electric field E ( r ) exists between the parallel plates (n.b. If the length × width (L × W) dimensions are both large in comparison to the plate separation, d i.e. L d and W d then the electric field E ( r ) between the parallel plates will be nearly uniform inside the gap region – i.e. E ( r ) E0 ( − zˆ ) = − E0 zˆ ← constant We know that ΔVab = Vb ( z = d ) − Va ( z = 0 ) = − ∫ E ( r )id and that b a of the path taken from a → b . ∫ E ( r )i d b a is independent So let’s simply take a path straight up along the ẑ axis − i.e. one that just barely touches the insides of the plates: Vb b +Q z=d we know: +++++++++++++++ E = E0 ( − zˆ ) = − E0 zˆ zˆ d =d d −−−−−−−−−−−−−−− Va a −Q z=0 both ΔVAb = Vb ( z = d ) − Va ( z = 0 ) = − ∫ E id = + Ed , for L, W a ) d both ΔVab Volts E= = constant for L, W m d 10 zˆ E = constant, ≠ fcn of ( z ) b ( ( + zˆ ) = + d d. ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede Then one can also see that if E = constant between conducting both ⎛ ⎞ plates ⎜ for L, W d ⎟ then ΔV ( z ) = Ez, where z is measured from the bottom plate (z = 0). ⎜ ⎟ ⎝ ⎠ From basic E&M principles, we know that E is linearly proportional to charge, Q ⎛ ⎞ 1 ⎛Q⎞ ⎜ cf : E ( r ) = ⎜ 2 ⎟ rˆ for a point charge ⎟ . 4πε o ⎝ r ⎠ ⎝ ⎠ ( Similarly, the potential, V is also linearly proportional to the charge, Q cf : ΔV = − ∫ E id c ) The constant of proportionality is known as the capacitance, C of the system, i.e. we define capacitance as the ratio of charge, Q to potential difference, ΔV associated with this system: C≡ Q ΔV S.I. units of capacitance is FARADS, F 1 Coulomb 1 FARAD = per volt pot'l difference both Thus, for the case of the parallel plate capacitor, for L, W capacitance of parallel plate capacitor ( L, W d) d we see that: Q Q C≡ = ΔVAB Ed Using Gauss’ Law (integral form) on (any) one of the parallel plates - e.g. the upper plate – use shrunken Gaussian pillbox as shown in figure below: ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. 11 UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede Then: ∫ s E idA = Qencl dA1 = dA1 zˆ, εo = ∫ E idA1 + S1 ∫ S2 E idA2 =0 E zˆ dA2 = dA2 ρˆ ( in x-y plane ) + ∫ E idA3 S3 E ⊥dA2 everywhere on S2 dA3 = − dA3 zˆ ( 0 zˆ )i( Adisk zˆ ) = =0 + ( − Ezˆ )i( − Adisk zˆ ) AS1 = AS3 = Adisk = π R 2 E = 0 inside upper plate (z > d ) Qencl = +σ Adisk = + EAdisk = Qencl εo =+ σ Adisk εo ∴ E= σ εo ∴ C= Q Q Q = = εo σd ΔVab Ed ∴ C= ε o Aplate d but Q = σ Aplate both for parallel plate capacitor, with L,W d ⇒ Note that capacitance, C depends only on ε o (electric permittivity of free space, 8.85×10−12 Farads/m) and (purely) geometrical factors (Aplate & d). 12 ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede Forces on a Parallel Plate Capacitor Suppose two parallel conducting plates, each of area A = L × W are separated by small distance d L,W . The potential difference between plates is initially ΔV . The top plate has surface charge density +σ , the bottom plate has surface charge density −σ . Top plate +σ plate separation distance d b Ο potential −σ ΔV difference a E= ΔV ΔV ( − zˆ ) = − zˆ d d ẑ ŷ x̂ Bottom plate ΔV = ΔVab = Vbtop − Vabottom Neglecting fringe fields effects @ edges of parallel plate capacitor (valid approximation if d L, W ): Pressure = force ⎛ Joules ⎞ = field energy density uE ⎜ ⎟ 3 per unit area ⎝ m ⎠ P= F 1 1 ⎛ ΔV ⎞ = uE = ε o E 2 = ε o ⎜ ⎟ A 2 2 ⎝ d ⎠ 2 2 1 1 ⎛ ΔV ⎞ Attractive force on (each) plate: F = P ⋅ A = uE ⋅ A = ε o E 2 A = ε o ⎜ ⎟ A 2 2 ⎝ d ⎠ (Newtons) Side view: ẑ +σ F top = − Fzˆ F bot = + Fzˆ E x̂ ŷ −σ ⎛ -plate ⎞ TOP BOT F ToT ⎜ = − Fzˆ + Fzˆ = 0 ⎟ =F +F ⎝ capacitor ⎠ No net force acts on parallel plate capacitor − Newton’s 1st Law: for every action, ∃ equal and opposite reaction. ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved. 13 UIUC Physics 435 EM Fields & Sources I Fall Semester, 2007 Lect. Notes 5 Prof. Steven Errede Griffiths EXAMPLE 2.11 Find the capacitance, C of two concentric spherical metal shells, with radii a & b, b > a. Place +Q on inner sphere and place –Q on outer sphere. Use Gauss’ Law and show that 1 ⎛Q⎞ for a ≤ r ≤ b : E ( r ) = ⎜ ⎟ rˆ 4πε o ⎝ r 2 ⎠ ⎧ b ⎪ΔVab = − ∫ E id a ⎪ b ⎪ Then: ⎨ = − ∫ E idr a ⎪ Q b 1 Q ⎛1 1⎞ ⎪ =− dr = ⎜ − ⎟ b>a 2 ∫ ⎪ a 4πε o r 4πε o ⎝ a b ⎠ ⎩ Then: Cconcentric ≡ spheres Q 1 ab = 4πε o = 4πε o ΔVab ⎛1 1⎞ (b − a ) ⎜ − ⎟ ⎝a b⎠ Again, note that C depends only on ε o and (purely) geometrical factors (radii a & b) How much work, W is done in charging up a capacitor − e.q. (if it is initially uncharged)? Charging an initially uncharged capacitor means individually removing electrons from the upper plate of the parallel-plate capacitor (inner sphere of concentric spherical capacitor) and transporting them to the lower plate of the parallel-plate capacitor (outer sphere of concentric spherical capacitor), respectively. W If ΔVab = (Griffiths 2.38, p.9) then W = QΔVab Q The infinitesimal amount of work dW needed to transport an infinitesimal amount of charge, dQ is dW = ΔVdQ But: ΔV = Q ⎛Q⎞ Therefore: dW = ⎜ ⎟ dQ C ⎝C ⎠ The total work done in charging a capacitor from Q = 0 to Q = QToT is: WToT = ∫ dW = ∫ Q = QToT Q =0 2 ⎞ 1 Q =QToT 1 ⎛ QToT ⎛Q⎞ = = dQ QdQ ⎜ ⎟ ⎜ ⎟ ∫ Q = 0 C 2⎝ C ⎠ ⎝C ⎠ constant 14 2 ⎞ 1 ⎛ QToT Thus: WToT = ⎜ ⎟ but: QToT = C ΔV 2⎝ C ⎠ 1 1 ⎛ Q2 ⎞ 1 ∴ WToT = C ΔV 2 = ⎜ ⎟ = Q ΔV 2 2⎝ C ⎠ 2 ©Professor Steven Errede, Department of Physics, University of Illinois at Urbana-Champaign, Illinois 2005 - 2008. All rights reserved.