- #1

- 1,030

- 4

Is there a simple way of proving that x=2^x can't happen for real x? Without series expansion.

You are using an out of date browser. It may not display this or other websites correctly.

You should upgrade or use an alternative browser.

You should upgrade or use an alternative browser.

- Thread starter Dragonfall
- Start date

- #1

- 1,030

- 4

Is there a simple way of proving that x=2^x can't happen for real x? Without series expansion.

- #2

arildno

Science Advisor

Homework Helper

Gold Member

Dearly Missed

- 10,025

- 135

Sure!

1. Note that if there were a solution, x cannot be negative, since the right-hand side is positve.

2. Note that at x=0, we have 2^x=1.

3. In order for "x" then, to become equal, (or greater) to 2^x, the slope of the function f(x)=x must be greater than the slope of the function g(x)=2^x prior to the x that makes f=g

4. Now, the slope of f is equal to 1, for g, it equals 2^xln(2)>0

5. Now, we see that g is strictly increasing.

IF the graph of f is to intersect somewhere with the graph of g, then the point at which g has equal slope as f must either be a point of intersection, OR IN A REGION IN WHICH f is greater than g!

(Since the graph of g will eventually undertake the graph of f, and if the point where g's slope is equal to f's slope is prior to f's intersection with g, then f will NEVER intersect with g)

6. Thus, we find that when g's slope equals f's slope, this happens at X=ln(1/ln(2))

7. At that point, we have f(X)=ln(1/ln(2)), whereas g(X)=(1/ln(2))^ln(2)

8 Since g(X)>f(X) we may conclude that g is always greater than f.

- #3

- 1,426

- 1

Much simpler is to write 2^x - x = 0, and take the derivative of LHS: ln(2)2^x - 1. On [0, ∞), this is clearly strictly positive, so the function 2^x - x is strictly increasing on that interval. At x = 0, 2^x - x = 1, so it never attains 0 on [0, ∞) (it only gets larger than 1). It also can never attain 0 on (-∞, 0), so there is no solution.

- #4

arildno

Science Advisor

Homework Helper

Gold Member

Dearly Missed

- 10,025

- 135

Most definitely not!Much simpler is to write 2^x - x = 0, and take the derivative of LHS: ln(2)2^x - 1. On [0, ∞), this is clearly strictly positive,

At x=0, for example, ln(2)-1 is negative.

- #5

- 1,426

- 1

Ah, then I retract my solution.

- #6

- 1,030

- 4

Ah, thanks for the replies!

- #7

uart

Science Advisor

- 2,797

- 21

This question could be turned into a nice little exercise for introductory calculus.

Q. Find the largest value of "

a = e^(1/e), with the solution occurring at x = e.

Share: