Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

My fragile attempt: Note that if $1987^k-1$ ends with 1987 zeros, that means $1987^k$ has last digit 1 (and 1986 "next" ones are zeros). For this to be satisfied, $k$ has to be in form $k=4n$, where $n\in N$. This means out number can be written in form

$$ [(1987^n)^2+1][1987^n+1][1987^n-1]. $$

This number has to be dividable by $10^{1987}$ if there is such a number that is asked for in question.

Now, I believe that the fact 1987 is a prime is very important here. There are probably some theorems from number theory about primes and their powers. For example, if $p$ is a prime (distinct from 2 if needed), are there any important things about number such as $p^2-1$?

If I'm going at the right direction with this, I'd appreciate a hint. Please don't use too advanced techniques if possible. Thanks.

share|cite|improve this question
So this is basically direct consequence of Carmichael function (I'm not familiar with Euler function so I looked up this generalization), since $GCF(1987,10^{1987})=1$? – Lazar Ljubenović Feb 7 '12 at 16:31
Yes, sort of. I'm not sure that one can know that the Carmichael function exists and makes sense without first knowing about the Totient function and Euler's theorem. – deinst Feb 7 '12 at 16:49
I don't, I just looked it up. This is beyond my level anyway, just wanted to make sure I at least understood general idea behind your comment. – Lazar Ljubenović Feb 7 '12 at 17:15
up vote 3 down vote accepted

The standard approach is to use the fact that if $a$ is divisible neither by $2$ nor by $5$, then $$a^{\varphi(10^n)}\equiv 1\pmod {10^n},$$ where $\varphi$ is the Euler $\varphi$-function.

The approach below is much more low-tech! All we need is some comfort with the Binomial Theorem. Suppose that $b_1$ already ends in $1$ (with our number, that means we let $b_1=(1987)^4$).
What happens when we take the $10$-th power of $b_1$?

Think of it this way. We have $b_1=1+10c_1$ for some integer $c_1$. Take (or imagine taking) the $10$-th power of $1+10c_1$, using the Binomial Theorem.

We get $1+(10)(10c_1)$ plus a bunch of terms that are divisible by at least $100$. So the result $b_2$ has shape $1+100c_2$, for some integer $c_2$. In other words, $b_2$ ends in $01$,

Now take the $10$-th power of $b_2$. We get, by the Binomial Theorem, $1+(10)(100c_2)$ plus a bunch of terms that are divisible by at least $1000$. Call this result $b_3$. Note that $b_3$ ends in $001$. Continue.

To sum up, we start with $(1987)^4$ and raise it to the power $10$ repeatedly. We get the numbers $(1987)^{40}$, $(1987)^{400}$, $(1987)^{4000}$, and so on. They are guaranteed to end in $01$, $001$, $0001$, and so on.

share|cite|improve this answer

Hint: You've already done the hard part, you have identified how the formula can be factored. Don't look for powers of 10, as all 3 terms are divisible by 2. Focus on powers of 5.

share|cite|improve this answer
Following your instructions, for now I could only prove divisibility by 5 for each modulo of $n$ by $4$. I'm not sure how to proceed and prove more powers of five. – Lazar Ljubenović Feb 7 '12 at 16:35
Yeah, you will need to look for the powers of 5. Try looking at the first thousand powers of 1987, modulus 100000. – Michael Sink Feb 7 '12 at 17:08
I don't know that skipping ahead to Euler is in spirit of the question, but if that works, great! There is a fairly simple formula for this particular problem. – Michael Sink Feb 7 '12 at 17:29

I don't think it's possible, here is why: We have established that in 1987^k k is of the form 4n or 2m and also that 1987^k ends in 1, so 1987^K = 1987^2m = (1987^m -1)(1987^m +1) Since 1987^k ends in 1, i.e. it is an odd number Both (1987^m -1) and (1987^m +1) are odd (as multiplication of only 2 odd numbers is odd) Now difference between 1987^m +1 and 1987^m -1 is 2 so basically we are multiplying 2 odd numbers with difference 2 and this multiplication ends with a number whose unit digit is 1. Now if we see all consecutive odd numbers (1, 3, 5, 7, 9 and again 1) it is not possible to have 1 as unit digit of the result of their multiplication. Hence 1987^k can not be number with unit digit 1 if k is of the form 2n hence 1987^K - 1 cannot end in number which ends at 0.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.