APPENDIX A Math Supplement 1015

areas of the individual rectangles in Figure A.8 in the limit within which the width of the
4
rectangles approaches zero. If the rectangles lie below the zero line, the incremental area
is negative; if the rectangles lie above the zero line, the incremental area is positive. In
this case the total area is zero because the total negative area equals the total positive area.
f(x)
This is the case because f(x) is an odd function of x. 2
The integral can also be understood as an antiderivative. From this point of view, the
integral symbol is defined by the relation
df 1x2 x
f1x2 = dx (A.39)
L dx
2 1 1 2
and the function that appears under the integral sign is called the integrand. Interpreting
the integral in terms of area, we evaluate a definite integral, and the interval over which
the integration occurs is specified. The interval is not specified for an indefinite integral. 2
The geometrical interpretation is often useful in obtaining an integral from experimental
data when the functional form of the integrand is not known. For our purposes, the interpre-
tation of the integral as an antiderivative is more useful. The value of the indefinite integral
1 1x - 5x2 dx is that function which, when differentiated, gives the integrand. Using the
3
4
rules for differentiation discussed earlier, you can verify that
x4 5x2 FIGURE A.8
1x3 - 5x2 dx = - + C (A.40)
L 4 2
Note the constant that appears in the evaluation of every indefinite integral. By differen-
tiating the function obtained upon integration, you should convince yourself that any con-
stant will lead to the same integrand. In contrast, a definite integral has no constant of
integration. If we evaluate the definite integral
2.3
x4 5x2 x4 5x2
1x3 - 5x2 dx = ¢ - + C≤ - ¢ - + C≤ (A.41)
3 4 2 x = 2.3 4 2 x = -2.3
-2.3
we see that the constant of integration cancels. Because the function obtained upon
integration is an even function of x, 1-2.31x3 - 5x2 dx = 0, just as we saw in the geo-
2.3

metric interpretation of the integral.

It is useful for the student of physical chemistry to commit the integrals listed next to
memory, because they are encountered frequently. These integrals are directly related to
the derivatives discussed in Section A.2:

df1x2 = f1x2 + C
L
xn + 1
xn dx = + C
L n + 1
dx
= ln x + C
L x
eax
eax = + C, where a is a constant
L a

sin x dx = -cos x + C
L

cos x dx = sin x + C
L
However, the primary tool for the physical chemist in evaluating integrals is a good set
of integral tables. The integrals that are most frequently used in elementary quantum
mechanics are listed here; the first group lists indefinite integrals:

1sin ax2 dx = -
1
cos ax + C
L a

1cos ax2 dx =
1
sin ax + C
L a
1016 APPENDIX A Math Supplement

1sin2 ax2 dx =
1 1
x - sin 2ax + C
L 2 4a

1cos2 ax2 dx =
1 1
x + sin 2ax + C
L 2 4a

1x2 sin2 ax2 dx =

1 3 1 1 1
x - ¢ x2 - 3 ≤ sin 2 ax - x cos 2ax + C
L 6 4a 8a 4a2

1x2 cos2 ax2 dx =

1 3 1 1 1
x + ¢ x2 - 3 ≤ sin 2 ax + x cos 2ax + C
L 6 4a 8a 4a2

1sin3 ax2dx = -
3 cos ax cos 3ax
+ + C
L 4a 12a
(a2x2 - 2) cos ax
x 1sin ax2dx =
2x sin ax
2
3
+ + C
L a a2
(a x - 2) sin ax
2 2
x2 1cos ax2dx =
2x cos ax
3
+ + C
L a a2
m ax
x e m
xmeax dx = - xm - 1eax dx + C
L a aL
eax 1 eax a eax
dx = - + dx + C
Lx
m
m - 1 xm - 1 m - 1 L xm - 1

The following group lists definite integrals.

a a

sin a b * sin a b dx = cos a b * cos a b dx = dmn

npx mpx npx mpx a
3 a a 3 a a 2
0 0
a

csin a b d * ccos a b d dx = 0
npx npx
3 a a
0
p p
p
sin mx dx =
2
cos2mx dx =
3 3 2
0 0
q q
sin x cos x p
dx = dx =
3 1x 3 1x A2
0 0

1a 7 0, n positive integer2
n!
xne-ax dx =
3 an + 1
0
q
1. 3 . 5 Á 12n - 12
1a 7 0, n positive integer2
p
x2ne-ax dx =
2
n+1 n Aa
3 2 a
0
q

1a 7 0, n positive integer2
n!
x2n + 1e-ax dx =
2

3 2 an + 1
0
q
p 1>2
e-ax dx = a b
2

3 4a
0

In the first integral above, dmn = 1 if m = n, and 0 if m Z n.

 APPENDIX A Math Supplement 1017

A.4.2 MULTIPLE INTEGRALS AND SPHERICAL COORDINATES

In the previous section, integration with respect to a single variable was discussed.
Often, however, integration occurs over two or three variables. For example, the wave
functions for the particle in a two-dimensional box are given by
nypy
cnxny1x, y2 = N sin
nxpx
sin (A.42)
a b
In normalizing a wave function, the integral of ƒ cnxny1x, y2 ƒ 2 is required to equal 1 over
the range 0 … x … a and 0 … y … b. This requires solving the double integral
b a
nypy 2
aN sin b dx = 1
nxpx
dy sin (A.43)
3 3 a b
0 0

to determine the normalization constant N. We sequentially integrate over the variables

x and y or vice versa using the list of indefinite integrals from the previous section.
b a
nypy 2
aN sin b dx
nxpx
dy sin
3 3 a b
0 0
b
2nxpx x = a nypy 2
= c x - d * N2 asin b dy
1 a
sin
2 4np a x=0 3 b
0
b
nypy 2
1 = c a - 1sin 2nxp - 02 d * N2 asin b dy
1 a
2 4np 3 b
0
N2ab
1 = N2 c a - 1sin 2nxp - 02 d * c b - 1sin 2nyp - 02 d =
1 a 1 a
2 4np 2 4np 4
2
N =
2ab
Convince yourself that the normalization constant for the wave functions of the three-
dimensional particle in the box
nypy nzpz
cnxnynz1x, y, z2 = N sin
nxpx
sin sin (A.44)
a b c
has the value N = 212> 1abc.
Up to this point, we have considered functions of a single variable. This restricts us
to dealing with a single spatial dimension. The extension to three independent variables
becomes important in describing three-dimensional systems. The three-dimensional sys-
tem of most importance to us is the atom. Closed-shell atoms are spherically symmetric,
so we might expect atomic wave functions to be best described by spherical coordinates.
Therefore, you should become familiar with integrations in this coordinate system. In
transforming from spherical coordinates r, u, and f to Cartesian coordinates x, y, and z,
the following relationships are used:
x = r sin u cos f
y = r sin u sin f (A.45)
z = r cos u
These relationships are depicted in Figure A.9. For small increments in the variables
r, u, and f, the volume element depicted in this figure is a rectangular solid of volume
dV = 1r sin u df21dr21r du2 = r2 sin u dr du df (A.46)
Note in particular that the volume element in spherical coordinates is not dr du df in
analogy with the volume element dxdydz in Cartesian coordinates.
1018 APPENDIX A Math Supplement

z rsin d
rd

rsin d dr
rd
rsin

dr
r d

FIGURE A.9

In transforming from Cartesian coordinates x, y, and z to the spherical coordinates

r, u, and f, these relationships are used:
z y
r = 2x2 + y2 + z2 u = cos-1 and f = tan-1 (A.47)
2x2 + y2 + z2 x
What is the appropriate range of variables to integrate over all space in spherical
coordinates? If we imagine the radius vector scanning over the range 0 … u … p;
0 … f … 2p, the whole angular space is scanned. If we combine this range of
u and f with 0 … r … q , all of the three-dimensional space is scanned. Note that
r = 2x2 + y2 + z 2 is always positive.
To illustrate the process of integration in spherical coordinates, we normalize the
function e-r cos u over the interval 0 … r … q ; 0 … u … p; 0 … f … 2p:
2p p q 2p p q

N2 df sin u du 1e-r cos u22r2 dr = N2 df cos2 u sin u du r2e-2rdr = 1

3 3 3 3 3 3
0 0 0 0 0 0

It is most convenient to integrate first over f, giving

p q

2pN2 cos2 u sin u du r2e-2r dr = 1

3 3
0 0

We next integrate over u, giving

q q
- cos3 p + cos3 0 4pN2
2pN c 2
d * r2e-2r dr = r2e-2r dr = 1
3 3 3 3
0 0

We finally integrate over r using the standard integral

q

1a 7 0, n positive integer2
n!
xne-ax dx =
3 an + 1
0
The result is
q
4pN2 4pN2 2! 3
r2e-2rdr = = 1 or N =
3 3 3 8 Ap
0

We conclude that the normalized wave function is 13>p e-r cos u.