Sunteți pe pagina 1din 6

Electric elds in steady-state superconductivity according to London theory

B. Nasmith and N. Gauthiera


Department of Physics, The Royal Military College of Canada, P. O. Box 17000, Station Forces, Kingston, ON K7K 7B4 Canada

Received 10 October 2007; accepted 21 April 2008 We show that a London superconductor in a steady uniform external magnetic eld must support an electric eld in its interior. The existence of an electric eld implies that a superconductor has a nonvanishing charge in its interior, a fact consistent with measurements of charge imbalance in steady-state superconductivity. We model the superconducting state as a collisionless zero-temperature electron gas. The London relation between the supercurrent velocity eld and the vector potential is used to show the existence of an induced electric eld. We argue that the inhomogeneities in the charge density give rise to a pressure gradient which we describe with the Fermi gas model. By analyzing the electric eld, we show that there are two length scales in the system the London penetration depth and a length associated with the compressibility of the electron gas. We determine the net electric potential in the interior of the superconductor. 2008
American Association of Physics Teachers.

DOI: 10.1119/1.2929671 I. INTRODUCTION The determination of the steady-state magnetic eld in the interior of a London superconductor immersed in a uniform external magnetic eld B0 has been discussed extensively.14 The reader will also nd comprehensive descriptions of the London theory in Refs. 510. Although the original macroscopic London theory was eventually supplanted by the microscopic BardeenCooper Schrieffer BCS theory, the former is still frequently used because it is more suited to describe certain macroscopic properties of superconductivity. The purpose of this paper is to examine some previously unexploited implications of the London theory that indicate that there are steady electric elds in the interior of superconductors. This conclusion is consistent with the results of numerous experiments of what is now referred to as charge imbalance in steady-state superconductivity since the original reports11,12 of this phenomenon in 1972. Experiments on charge imbalance1315 have been analyzed within the framework of BCS theory. The present approach might provide a complementary viewpoint and help to develop a more complete understanding of the phenomenon. In the London theory the form of the equations that govern the ow of supercurrents is justied in large part by the fact that they correctly describe the steady-state Meissner effect.16 This effect causes the magnetic eld to be expelled from the interior of the superconductor, only to persist in a thin surface layer whose thickness is equal to a few times the London penetration depth L, which is of order 15 nm, and is given by:
L

The basic assumption of the London theory for steadystate superconductivity is the constitutive relation between the velocity eld of the supercurrents v = v r and the vector potential A = A r at point r in the superconductor: v= e A. m 2

This basic relation leads, using Maxwells fourth equation, to the second-order partial differential equation for the magnetic vector potential: 1
2

A= 0
2

2A L

inside the superconductor outside the superconductor,

m n 0e 2

.
0

In Eq. 1 m = 9.1x1031 kg and e = 1.6x1019 C are the mass and absolute value of the electric charge of an electron, respectively, and 0 = 4 x107 Hm1 is the magnetic permeability of vacuum. Also, n0 is the uniform electron number density per unit volume in the absence of an externally applied magnetic eld.
872 Am. J. Phys. 76 9 , September 2008 http://aapt.org/ajp

with the Laplacian operator. The magnetic eld B is related to the vector potential A by B = A so by taking the curl of Eq. 3 we see that the magnetic eld also obeys an equation of the same form. Discussions of the steady-state magnetic properties of London superconductors are common, but their electric properties are seldom mentioned. Fritz London wondered if a steady electric eld could exist in the interior of a superconductor, but ruled out that possibility by stating: for stationary conditions there is no electric eld in the superconductor.10 Italics are in the original. Given the experimental evidence that contradicts this statement,1115 we revisit the issue and examine the implications of the London theory for Newtons law of motion for the supercurrents. We restrict our analysis to practical situations where the linear dimensions of the superconductor are much larger than the London penetration depth, a restriction that will allow us to simplify the results. In this paper we show that there is an induced charge in the interior of a superconductor immersed in a steady magnetic eld. The presence of this induced charge indicates that the electron gas is put under pressure by the magnetic eld. We will also show that a negative layer of polarization charge is exposed on the surface of the superconductor. Our analysis suggests that the penetration depth associated with
2008 American Association of Physics Teachers 872

the Fermi pressure is much smaller than the London penetration depth; we call the former the Fermi penetration depth and denote it by F. We will show that F 0.05 nm for typical London superconductors. Our analysis will be applied to the case of a spherical superconductor and the behavior of the solutions will be discussed.

E=

m v e

vv

1 ene

pF .

II. THE LONDON CONSTITUTIVE RELATION AND THE LORENTZ AND FERMI PRESSURE FORCES We rst assume that the superconductor is immersed in a uniform external magnetic eld. Then we use the London constitutive relation of Eq. 2 and determine its implications for the motion of an element of supercurrent within the zerotemperature Fermi gas model.4,17 To reect the presence of local charge imbalances that create the electric eld we write the position-dependent superelectron number density ne = ne r as ne n 0 + n 0u r , 4

The term v v on the right-hand side of Eq. 9 represents the acceleration of the stationary closed loops formed by the supercurrents. If we use cylindrical coordinates with v = v , z for an axially symmetric superconductor, we nd that this term equals v2 / . This term is the centripetal acceleration of the supercurrents at a point, and is directed toward the axis about which the supercurrents are rotating, the z axis. We now invoke the London constitutive relation of Eq. 2 to express the velocity eld in terms of the vector potential in Eq. 9 to nd an expression for the electric potential. We rst note the identity18 A A+A A = 1 2 A . 2 10

Then we substitute Eqs. 2 and 6 into Eq. 9 , apply Eq. 10 and replace ne by n0 in front of pF to obtain the electric eld to linear terms in u: E=
0 eA2 2EF + u 2m 3e

where n0 is the unperturbed or zero-eld electron number density and u = u r is a dimensionless quantity that measures its relative deviation at r , u = ne n0 / n0. A mass m and a charge -e are associated with each moving particle so that the mass density of the moving electrons per unit volume is equal to mne. If we insert the Lorentz force density and the Fermi pressure gradient in the right-hand side of Newtons second law for the motion of an element of supercurrent, we nd d mne v = ene E + v dt B pF . 5

11

In Eq. 11 , is the net electric potential within the superconductor. We take the potential contributions to vanish in zero applied magnetic eld and write: =
b

s,

0 eA2 2EF + u b, 2m 3e

and

12

0 2EF us . 3e

In Eq. 5 E is the net electric eld in the superconductor and pF is the Fermi pressure, which is, in the degenerate Fermi gas model,4,17 2 2 0 p F = E Fn e = E Fn 0 1 + u 5 5
5/3

0 0 2n0EF 2n0EF + u+ . 5 3

In Eq. 6 EF is the Fermi energy in the presence of an applied magnetic eld,


2

In Eq. 12 we have separated the net electric potential into a bulk contribution b and a surface contribution s. We have also separated the net relative deviation of the electron number density u = ne n0 / n0 into a bulk contribution ub and a surface contribution us: u = ub + us. As we shall see both contributions extend into the interior of the superconductor but are characterized by very different penetration depths. The net charge density at a point is, with Eq. 4 :
net =

en0ub en0us .

13

EF =

2m

ne

2/3

0 = EF 1 + u

2/3

0 2 EF 3 2n0 2/3 / 2m is the unperturbed or zero-eld Fermi energy, and is Plancks constant divided by 2 . Equation 4 was used for ne and the right-hand side of Eq. 6 was expanded to linear terms in the small quantity u. The pressure gradient term in Eq. 5 is analogous to a similar term in the Bernoulli theory for the motion of a nonviscous uid and is included here to account for the fact that ne is nonuniform. Finally, d / dt is the co-moving time derivative and v = v r is the steady supercurrent velocity eld. We have, by the chain rule, that:

We will show that ub is positive and that us is negative. If we take the divergence of the net electric eld in Eq. 11 , invoke Maxwells rst law, and rearrange terms, we nd that ub + us =
0 2 eA 2

en0

2m

2 F

ub +

2 F

us ,

14

where 0 = 8.85x1012 Fm1 is the permittivity of free space. The quantity F has units of length and will be called the Fermi penetration depth. This new penetration depth is associated with the compressibility of the Fermi gas 0 = 3 5EFn0 1 and is given by
F 0 2 0E F . 3n0e2

d v= v+ v dt t

v= v

v.

15

We now use Eq. 8 to solve Eq. 5 for the net electric eld and nd
873 Am. J. Phys., Vol. 76, No. 9, September 2008

0 Using typical values of EF = 10 eV and n0 = 15x1028 m3 we 12 2 1 get 2.5x10 m N .

B. Nasmith and N. Gauthier

873

We now require without any loss of generality that ub satisfy the equation: ub =
0 2

we integrate Gauss ux law on the surface of the supercon0 ductor and obtain, with Eb b , that Qb =
0

en0

0 b

2 F

ub ,

16

ds Eb =

e 0 2m

ds A2

0.

21

where b0 eA2 / 2m . We assume that the vector potential is known from the solution of Eq. 3 . There is no exibility left for the terms that contain us in Eq. 14 and us must satisfy the second-order differential equation: us =
2 F 2

In the surface integrals in Eq. 21 , ds is the outwardoriented element of surface area. A distributed charge Qs is also present and is given by integrating the second term on the right-hand side of Eq. 13 over the entire superconductor: Qs = d en0us . 22

us .

17

Iterations of Eq. 16 are straightforward and give the series: ub = =


0 2

en0
0 2

0 b

+ +

2 F 0

en0
2 F 4 0 b

0 b

+
0

2 F

ub
0 b

en0

0 b

en0

en0

4 F

+ 18

In this expression, d is the element of volume. We will show in the next section that the net charge in the interior of the superconductor, Qnet = Qb + Qs, is positive, which implies, through the overall electrical neutrality of the system, that there is a negative layer of polarization charge on the surface of the superconductor. III. APPLICATION TO THE SPHERICAL SUPERCONDUCTOR We now consider a spherical superconductor and examine some of the implications of the present theory. Assume that a spherical superconductor of radius R is immersed in a uni form z-directed applied magnetic eld B0 = kB0. The induced supercurrents form closed circular loops centered on the z axis whose plane is normal to that axis. We dene the dimensionless radial coordinates w r / L and W R / L. Spherical coordinates are used. The components of the magnetic eld in the interior of the superconductor are1 Br = b0 w cosh w sinh w cos , w3 23

ub0 + ub2 + ub4 + .

The terms in Eq. 18 can all be evaluated in principle because b0 can be found by solving Eq. 3 . We now argue that this series is characterized by the small parameter 2 F / L . This claim is supported by our results for the spherical superconductor of radius R in Sec. III because in this case the potential b0 behaves asymptotically for 1 r / L R / L as e2 rR / L, where r is the distance from the center of the sphere. To determine the convergence of the series in Eq. 18 we need to estimate L and F for typical London superconductors such as lead and indium. The experimental values4 for these materials are n0 = 13.2x1028m3, 0 0 EF = 9.37 eV for Pb, and n0 = 11.49x1028 m3, EF = 8.60 eV for In. As a result, we nd that L Pb = 14.6 nm, F Pb = 51.2 pm, and L In = 15.7 nm, F In = 52.5 pm. For these two cases F / L 2 105, which means that we can neglect all the terms in the last line of Eq. 18 , except the rst one. We then write that: ub ub0 =
0 2

B =

b0 w cosh w 1 + w2 sinh w sin , 2 w3

24

en0

0 b

19

We now return to Eq. 17 and note that the small value of validates our suggestion that us describes a phenomenon that is localized close to the surface of the superconductor. Indeed we have found that F is roughly equal to the radius of a hydrogen atom. Accordingly, due to Eq. 17 , us decays exponentially below the surface of the superconductor within a distance of F. The very stiff electron gas is disturbed only within a layer of thickness equal to a few times F near the surface. We shall show in Sec. III that the associated static electric eld and potential are prohibited from penetrating too deeply into the superconductor. We now examine the charge density at points in the bulk of the superconductor. The bulk charge density is given by the rst term in the right-hand side of Eq. 13 and we have, from Eq. 19 , that:
F b

with b0 3B0W / sinh W . The associated vector potential satises Eq. 3 , has the symmetry-imposed form A = A r , , and may be written as follows: A = a0i1 w sin i1 w = ; 0 w W 25 26

w cosh w sinh w . w2

The constant a0 3 LB0W / 2 sinh W has the units of a vector potential, and i1 w is a dimensionless function, regular at the origin, that obeys the modied spherical Bessel equation of order k = 1:20 ik w + 2 k k+1 ik w 1 + w w2 ik w = 0. 27

en0ub =

e 0 2m

A2

0.

20

The bulk charge density is negative because the vector potential decays exponentially below the surface, as will be conrmed in Sec. III. To obtain the total bulk charge Qb,
874 Am. J. Phys., Vol. 76, No. 9, September 2008

The prime indicates a derivative with respect to the functions argument. We now consider the relative deviation of the bulk number density ub of Eq. 19 and determine its contribution to the electric potential. We rst use Eqs. 25 and 26 and nd that
B. Nasmith and N. Gauthier 874

0 b

e a0i1 w sin 2m

= Hi2 w P0 Hi2 w P2 , 1 1

28

a0

3B0ReW

and

3eB2R2 2W 0 e . m

34

where H = ea2 / 3m. We write Pk for the Legendre polynomial 0 1 of order k.19,20 Because P0 = 1 and P2 = 2 3 cos2 1 , we can write sin2 = 2 P0 P2 / 3. We then nd from Eq. 19 that the bulk contribution to the relative deviation of the number density is ub = = =
0H 2 2 i1

Finally we let W w = R r / L be the distance below the surface of the superconductor relative to the London penetration depth and express the asymptotic behavior of the vector potential, Eq. 25 , of the motion-induced scalar potential, Eq. 28 , and of the normalized deviation of the number density in the bulk, Eq. 29 , as follows: A 3B0 2
0 b L

en0
0H 2 en0 L 0H 2 en0 L

w P0 P2 w2 i2 w 1 + i2 w 1 w w P0 w 6i2 1 w w
2 2

e sin ,
2 L

35

1 w2

P0 P2

9eB2 0 8m

e2 sin2 ,

36

w2 i2 w 1 1 w w2 w w 1 2 w w
2

ub P2 29

0H 2 en0 L

i2 1 w

9 0B2 2 0 e sin2 . 2mn0

37

g0 P0 + g2 P2 . To obtain Eq. 29 we used the relation19,20


2 2

Pk

+ cot

Pk = k k + 1 Pk .

30

If we calculate the derivatives in Eq. 29 and use Eq. 27 with k = 1 to eliminate i1 w , we nd: g0 w = =
0H 2 en0 L

These three quantities decay exponentially from the surface of the superconductor toward its interior. Equations 29 , 31 , and 32 were used to obtain the asymptotic behavior in Eq. 37 . Note that Eq. 36 supports our claim regarding the rapid convergence of the perturbation series in Eq. 18 . We now evaluate the potential contribution s 0 = 2EF / 3e us, which satises Eq. 17 because it is proportional to us, and then calculate the net internal electric potential. Let us dene the dimensionless radial variable, s r / F, with S R / F. Equation 17 then gives the following differential equation for s: 1 s2 s s2 s +
2 s

2i1

2 w + 2i1 w i1 w + i1 w w 2 2 i w ; w2 1 2 i1 w w 31

= 0.

38

2 0H en0
2 L

i12 w + 1 +

The angular part of the Laplacian is as dened in Eq. 30 . The solution of Eq. 38 that is regular at the origin and has angular terms in P0 and P2, as in the forcing term of Eq. 28 , may be written as follows:
s

g2 w = =

0H 2 en0 L

2i12 w + 2i1 w i1 w +

i0 s i2 s P0 + P2 . i0 S i2 S

39

6 2 i w w2 1 2 0H
2 en0 L

i12 w + 1

1 2 i w w2 1

The constants and are determined in the following. The functions i0 s and i2 s satisfy the modied spherical Bessel equation20 of Eq. 27 , with k = 0 and k = 2, respectively. These two functions and their asymptotic behavior are . 32 i0 s = sinh s s es , 2s es . 2s 40

As mentioned, g0 and g2 decrease exponentially over a distance characterized by the London penetration depth as w moves from the surface into the interior of the superconductor. We may therefore replace i1 w and its rst derivative i1 w by their asymptotic expansions in Eqs. 31 and 32 . We nd from Eq. 26 and from its rst derivative, that i1 w ew , 2w i1 w ew , 2w 1 w W . 33

3 3 1 i2 s = 3 + sinh s 2 cosh s s s s

For the same reasons that we gave after Eq. 33 , we now replace s by S in the denominators in Eq. 40 and replace both i0 S and i2 S by eS / 2S, to get:
s

P0 + P2 =

3 e sin2 . 2

41

We note that 1 / w varies very slowly with w compared to ew in Eq. 33 , and we replace the former by 1 / W because 1 w W. We also replace sinh W by eW / 2 in the expressions for a0 and H in Eqs. 25 and 28 because W 1 and get that:
875 Am. J. Phys., Vol. 76, No. 9, September 2008

S s = R r / F is the distance below the In Eq. 41 surface of the sphere, normalized to the Fermi penetration depth. We have also used the angular symmetry of the forcing term in Eq. 28 and of the bulk contribution ub in Eq. 37 to write that = . To determine the constant we rst
B. Nasmith and N. Gauthier 875

note that the potential outside the superconductor, , satises the Laplace equation and the solution for the present situation is d0 S / s P0 + d2 S / s 3 P2, where d0 and d2 are constants associated with the monopolar and the quadruploar terms of the charge distribution, respectively. Furthermore, we must have that d2 = d0 because the surface potential is continuous and must then have the symmetry shown by the net interior potential , which is the sum of Eqs. 36 and 41 : = b0 + s. We neglected the potential contribution 0 2EF / 3e ub in this expression because it is of order F / L 2 compared to b0 . Finally the overall electrical neutrality of the superconductor requires that there be no monopolar contribution to the exterior potential, which means that d0 = 0, d2 = 0. The electric potential therefore vanishes at the surface of the superconductor and this condition allows us to determine the unknown constant of Eq. 41 . Adding together Eqs. 36 and 41 and setting the result equal to zero at the surface of the superconductor then gives that 2 = 3eB2 L / 4m . We are now in a position to calculate the 0

net charge that resides in the interior of the superconductor, as discussed in the nal paragraph of Sec. II. Indeed, we must have, from Eqs. 37 , 41 and 12 that
R 2

Qnet =
0

drr2
0

d sin
0 R

en0 ub + us
L

= 2 en0
0

drr2 e Rr /

9 0B2 2 Rr / 0 e 2mn0
F

9 0B 2 0 2 3

2 L 2 8mn0 F

d sin P0 P2
0

42

To get the last line above, we used sin2 = 2 P0 P2 / 3. We now note that the polar integration involving P2 vanishes while the one involving P0 equals 2. The radial integral is easily done using the variable , and we nd that

Qnet = 4 en0 4 e 4 R2 =4 R2

3 0B 2 0 mn0 3
2 0B 0

R/ F

d R
0

e2

F / L

2 L 2e F

m
F

R2

R/ F

d
0 L

e2
2 L 2 F

F / L

2 L 2e F

3 0eB2 0 m 3
2 0eB0 L

+
L

2m

1+

0.

43

To get the second line in Eq. 43 we neglected the term F in R F 2 because of the rapid decay of the exponential factors that multiply it. We also neglected terms in exp R / F and exp 2R / L in the third line because R F, L. There is therefore a negative polarization charge on the surface of the superconductor and it is characterized = 0 P0 P2 , where by a surface density

0 b

+ 8m

2 9eB2 L 0

sin2 ,

45

where f e2
F / L

e .

46

3 0eB2 0 0= 2m 3 4m

1+ .

2 2 0eB0 L F

44

Finally we determine the net electric potential inside the superconductor. It is worth mentioning here that the boundary condition for the surface discontinuity in the radial component of the electric eld leads us exactly to the same conclusion but we omit the details for the sake of brevity. The asymptotic expression for the net electric potential at points within the superconductor is:
876 Am. J. Phys., Vol. 76, No. 9, September 2008

Note that in the region of interest here, is positive and increases from the surface of the superconductor = 0 toward its center. for F = 0.05 nm and L = 15 nm, Figure 1 is a plot of f 20. This range that is, for L / 2 F = 150 in the range 0 extends below the surface of the superconductor to 20 Fermi vanishes at the surface penetration depths. The function and at the center of the sphere and it goes through an extremum at 5. We now estimate the net electric potential inside the superconductor using the values B0 = 0.1 T and 9 m. We nd that = + 0.4f sin2 V. At L = 15x10 5, f is maximum and equal to 0.96 so that the peak V. This completes our discuspotential equals + 0.4 sin2 sion of the induced electric potential in the interior of the superconductor.
B. Nasmith and N. Gauthier 876

Fig. 1. Plot of f = e2 F/ L e vertical versus R r / F horizon20 for L / 2 F = 150. This function is the superposital in the range 0 tion of two decaying exponentials with very different penetration depths. One horizontal unit corresponds to one Fermi penetration depth; 300 units correspond to one London penetration depth. The function vanishes at = 0, then rises and reaches a positive maximum around = 5 and decays back to zero at the sphere center.

IV. CONCLUDING REMARKS We have argued that there is a conservative electric eld in the interior of a London superconductor. This eld must satisfy Newtons second law of motion for the supercurrents because the closed supercurrent loops require the action of a net force that is not provided by the magnetic eld alone. We also showed that the applied magnetic eld exerts pressure on the electron gas and this additional force was modeled using the Fermi gas theory. Finally we argued that the gradient of the Fermi pressure gives rise to a new penetration depth that is roughly 300 times smaller than the London penetration depth. ACKNOWLEDGMENTS The authors wish to thank the two referees for very valuable suggestions. In particular, we owe the idea of introducing the Fermi penetration depth as shown in Eq. 14 to one of the referees. We also thank Dr. J. Buckley for preparing the gure. This work is part of B. Nasmiths research project for a M.Sc. degree in physics.

Electronic mail: gauthier-n@rmc.ca E. Matute, On the superconducting sphere in an external magnetic eld, Am. J. Phys. 67, 786788 1999 . 2 J. R. Reitz, F. J. Milford, and R. W. Christy, Foundations of Electromagnetic Theory AddisonWesley, Reading, MA, 1979 , 3rd ed. 3 V. V. Batygin and I. N. Toptygin, Problems in Electrodynamics Academic, London, 1978 , 2nd ed. 4 C. Kittel, Introduction to Solid State Physics Wiley, New York, 1976 , 5th ed. 5 C. P. Poole, Jr., H. A. Farach, and R. J. Creswick, Superconductivity Academic, New York, 1995 . 6 W. Buckel, Superconductivity VCH Verlagsgesellschaft, New York, 1991 . 7 M. Crisan, Theory of Superconductivity World Scientic, London, 1989 . 8 P. G. de Gennes, Superconductivity of Metals and Alloys W. A. Benjamin, New York, 1966 . 9 F. London and H. London, The electromagnetic equations of the supraconductor, Proc. R. Soc. London, Ser. A 149, 7188 1935 . 10 F. London, Superuids Wiley, New York, 1950 , Vol. I, p. 33. 11 J. Clarke, Experimental observation of pair-quasiparticle potential difference in nonequilibrium superconductors, Phys. Rev. Lett. 28, 1363 1366 1972 . 12 M. Tinkham and J. Clarke, Theory of pair-quasiparticle potential difference in nonequilibrium superconductors, Phys. Rev. Lett. 28, 1366 1369 1972 . 13 H. J. Mamin, J. Clarke, and D. J. Van Harlingen, Charge imbalance induced by a temperature gradient in superconducting aluminum, Phys. Rev. B 29, 38813890 1984 . 14 J. E. Hirsch, Consequences of charge imbalance in superconductors within the theory of hole superconductivity, Phys. Lett. A 281, 4447 2001 . 15 R. Nicolsky and S. Frota-Pessa, Voltage in a charge-imbalance experiment, taking into account the approximate charge neutrality in the superconductor, J. Low Temp. Phys. 46, 107113 1982 . 16 W. Meissner and R. Ochsenfeld, Ein neuer Effekt bei Eintritt der Supraleitfhigkeit, Naturwiss. 21, 787788 1933 . English translation available in: A. M. Forrest, Meissner and Ochsenfeld revisited, Eur. J. Phys. 4, 117120 1983 . 17 P. M. Morse, Thermal Physics Benjamin/Cummings, Reading MA, 1978 , 2nd ed. 18 P. Lorrain, D. P. Corson, and F. Lorrain, Electromagnetic Fields and Waves W. H. Freeman, New York, 1988 , 3rd ed. 19 R. Haberman, Elementary Applied Partial Differential Equations PrenticeHall, Englewood Cliffs, NJ, 1983 . 20 M. Abramowitz and I. A. Stegun, editors, Handbook of Mathematical Functions Dover, New York, 1972 .
1

BUT THEN YOU PROBABLY KNEW THAT Observers worry that poor science education will lead to decreased support for science and fewer students seeking scientic careers; in fact, the current U.S. production of college graduates in science and engineering trails the growth of jobs in these areas, with the difference made up by people from abroad. Most important, the failure to convey science to the populace means we lack a well-informed U.S. citizenry when it comes to the scientic issues confronting us all, from stem-cell research to hurricane prediction and nuclear power.
Sidney Perkowitz, Hollywood Science: Movies, Science, and the End of the World Columbia University Press, 2007 , p. 217.

877

Am. J. Phys., Vol. 76, No. 9, September 2008

B. Nasmith and N. Gauthier

877

S-ar putea să vă placă și