Math 319 Problem Set 3: Complex numbers and Quaternions Lie Groups This problem set introduces two number systems that will play important roles in our work this semester, the complex numbers C and the quaternions H. (For some history, see the photocopied excerpt from Conway and Smith.) It also gives more practice with the definition of a group. We begin with the complex numbers, which will give us another way to think about the groups O(2, R) and SO(2, R) from Problem Set 1. C = {a + bi : a, b ∈ R}, where addition and multiplication are defined by i2 = −1 and the usual rules in R. That is, (a + bi) + (a0 + b0 i) = (a + a0 ) + (b + b0 )i (a + bi)(a0 + b0 i) = (aa0 − bb0 ) + (ab0 + ba0 )i. You can assume that addition and multiplication are associative and commutative and satisfy the distributive law (straightforward to verify). Another “operation” on C is complex conjugation, where the complex conjugate of z = a + bi is z = a − bi. Notice that z = z if and only if z is a real number. Also, for any z, zz = a2 + b2 is a real number, and zz is nonzero if z is nonzero. Also notice that nonzero complex numbers have multiplicative inverses, z −1 satisfying zz −1 = z −1 z = 1, namely µ ¶ 1 a 1 b −1 z= 2 z = = − i. z zz a + b2 a2 + b2 1. Verify that z = z, z1 + z2 = z 1 + z 2 , z1 z2 = z 1 z 2 , and 1/z = 1/z. 2. a. Show the complex numbers form a group under addition. You have to check three things: • Show that C is closed under addition — the sum of two complex numbers is again a complex number. • Show that there is an identity for addition in C. [Explain why 0 is the right choice for the identity for addition, i.e. the analog of I for composition of transformations or In for multiplication of matrices.] • Show that every complex number has an additive inverse which is again a complex number. [What is the inverse of a complex number for addition?] 1 b. Show the nonzero complex numbers form a group under multiplication. You have to check three things: • Show that C is closed under multiplication — the product of two nonzero complex numbers is again a nonzero complex number. • Show that there is an identity for multiplication among the nonzero complex numbers. [Explain why 1 is the right choice for the identity for multiplication, i.e. the analog of I for composition of transformations or In for multiplication of matrices.] • Show that every nonzero complex number has a multiplicative inverse which is again a nonzero complex number. [See the introduction to this problem set.] A standard way to draw a picture of a complex number is to represent a + bi by the point (or, equivalently, the vector) (a, b) ∈ R2 . In this representation, the horizontal axis is the real axis, and the vertical is imaginary. The length of the vector corre√ sponding to z = a + bi is then zz. Notice that a complex number of “length” 1 is thus one of the form cos(α) + i sin(α) for some real number α. 3. Recall that the Taylor series (centered at 0) for the exponential function is ex = 1 + x + x2 xn + ··· + + ··· 2! n! and for the sine and cosine sin(x) = x − x3 x5 x2m+1 + − · · · + (−1)m + ··· 3! 5! (2m + 1)! and x2m x2 x4 + − · · · + (−1)m + ··· 2! 4! (2m)! a. Substitute x = iα into the series for ex and, comparing to the series for sine and cosine, show that eiα = cos(α) + i sin(α). b. Use (a) to show e0 = e2πi = 1. What is eπi ? c. Use the rules for multiplying complex numbers and trig identities to verify that eiα eiβ = eiα+iβ = ei(α+β) . cos(x) = 1 − 4. In a natural way, each complex number eiα can be identified with a point of the unit circle S 1 in R2 . [The unit circle is sometimes called a 1-sphere — hence the exponent 1 in S 1 — because it is 1-dimensional, a curve.] Similarly, each point of the unit circle can be identified with a complex number of the form eiα : S 1 = {eiα : 0 ≤ α < 2π} = {z ∈ C : |z| = 1} = {z ∈ C : zz = 1} a. Use the description you prefer (and previous results, as needed) to show that S 1 is a group under the usual multiplication of complex numbers. That is, show 2 • S 1 is closed under multiplication: if z1 and z2 belong to S 1 , then so does their product z1 z2 . • The complex number 1 is in S 1 . • If z is in S 1 then its multiplicative inverse z −1 is also in S 1 . 5. Fix the real number α. Define a function Tα : C → C by Tα (z) = eiα z. a. Show that Tα preserves “lengths” of complex numbers. That is, assume w = Tα (z), and then show ww = z z. (Such a transformation is an isometry of C.) b. Interpreting z as a vector in R2 , show that Tα is a linear transformation. That is, show Tα (z1 + z2 ) = Tα (z1 ) + Tα (z2 ) and Tα (rz) = rTα (z) for z1 , z2 , z ∈ C and r ∈ R. c. What is the matrix of Tα with respect to the standard basis of R2 ? 6. Again fix the real number α. This time define a function Sα : C → C by Sα (z) = eiα z. a. Show that Sα preserves lengths of complex numbers. b. Interpreting z as a vector in R2 , show that Sα is a linear transformation. c. What is the matrix of Sα with respect to the standard basis of R2 ? 7. We’ll also have occasion to use matrices with complex entries: M(n, C) is the set of all n by n matrices with entries in C. If the matrix A ∈ M(n, C) has entries aij , write A for the matrix with entries aij . (Similarly, the complex conjugate of a vector v = (v1 , . . . , vn ) is v = (v1 , . . . , vn ).) Complex conjugation has a simple relationship to matrix operations on A, B ∈ M(n, C). For each of the following, you may make your argument for the case when n = 2, but to make the generalization to arbitrary n easier later, write your two by two matrices in the form · ¸ a11 a12 A= a21 a22 a. b. c. d. e. Show A + B = A + B. Show that AB = A B. Show AT = (A)T . Show tr(A) = tr(A). Show det(A) = det (A). 3 8. This is getting a bit ahead of our story, but there’s a very important group of matrices with complex entries called the unitary group U (n, C). T U (n, C) = {M ∈ M(n, C) : M M = In } Assume the results of (7) for the following. a. Show that U (n, C) is a group under matrix multiplication. b. Show that SU (n, C) is a subgroup of U (n, C), where SU (n, C) = {M ∈ U (n, C) : det(M ) = 1 }. c*. Show that if M ∈ SU (2, C), then · M= a −b b a ¸ for complex numbers a and b satisfying aa + bb = 1. [Hint: Set up notation · ¸ a c M= , b d and write out the equations involving a, b, c, d that correspond to the requirements T M M = I2 and det(M ) = 1. Use the equations to write c and c in terms of a and b, but consider two cases separately: Case (i) a = 0, and Case (ii) a 6= 0.] Remark. In Part 5 we’ll motivate this definition more fully and do more with it. Looking ahead, physicists use four groups to make predictions about the “electron shells” and “energy levels” of a hydrogen atom: SO(3, R), SO(4, R) and SU (2, C). (See Linearity, Symmetry, and Prediction in the Hydrogen Atom by Singer.) 9*. The quaternions H (H is for Hamilton — see Conway and Smith and also Yaglom) are a generalization of the complex numbers: H = {a + bi + cj + dk : a, b, c, d ∈ R}, where i2 = j 2 = k 2 = −1, ij = k = −ji, jk = i = −kj, and ki = j = −ik. Otherwise, addition and multiplication follow the usual rules for R: for z = a + bi + cj + dk and z 0 = a0 + b0 i + c0 j + d0 k, z + z 0 = (a + a0 ) + (b + b0 )i + (c + c0 )j + (d + d0 )k zz 0 = (aa0 − bb0 − cc0 − dd0 ) + (ab0 + ba0 + cd0 − dc0 )i +(ac0 + ca0 + da0 − ad0 )j + (ad0 + da0 + bc0 − cb0 )k Addition and multiplication are associative and satisfy the distributive law. Addition is commutative, but multiplication is not. An analogue of complex conjugation is also defined for H: z = a − bi − bj − bk. 4 a. Show z + z 0 = z + z 0 and z = z for all z, z 0 ∈ H. b. Calculate zz and find a formula for the multiplicative inverse 1/z for z ∈ H, z 6= 0. [Hint: imitate the complex case.] c. Show H is a group under addition. d. Show the nonzero elements of H form a group under multiplication. e. Since multiplication in H isn’t commutative, perhaps it’s not surprising that the relationship between multiplication and complex conjugation is different for H than for C. Show zz 0 = z 0 z for all z, z 0 ∈ H. [Hint: you can simplify an otherwise messy calculation by writing z = a + w for w = bi + cj + dk and observing z = a − w. Then you just have to check that w0 w = ww0 .] f. Show {z ∈ H : zz = 1} is a group under multiplication ( so it is a subgroup of the multiplicative group of nonzero quaternions). Remark: When z ∈ H is identified with the point (a, b, c, d) ∈ R4 , the group in (f) is often called the 3-sphere S 3 in R4 , by analogy to the 1-sphere S 1 in R2 . The set S 2 is the surface of a “regular” unit sphere in R3 ; it does not have a natural group structure. 5