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

Let $a,b,c,d$ be natural numbers with $ab=cd$. Prove that $a+b+c+d$ is composite.
I have my own solution for this (As posted) and i want to see if there is any other good proofs.

share|cite|improve this question
Is this homework? What have you tried? – Alex Provost May 6 '13 at 15:38
I vote against closing this question. – MJD May 6 '13 at 15:45
Let the OP respond before closing the question. The question seems pretty good. – Inceptio May 6 '13 at 15:48
I, too, find this question interesting. Nevertheless I downvoted it for the way it was asked: No context was given, no own effort is shown and only in the comments it turns out it was merely asked to see if someone will come up with a better solution and that the questioner already has solutions. So my read on it is that it’s basically a test to see whether other people are as clever as the questioner which I find rude. But maybe I misinterpret this? At the very least, I think you, CODE, should add your intentions to the question. – k.stm May 6 '13 at 16:10
Note this does not work if you include 0 in the natural numbers. (Obvious counterexample: a = 0, b=3, c=0, d=4, so ab = cd = 0, but a+b+c+d=7 which is prime). – dr jimbob May 6 '13 at 18:49
up vote 23 down vote accepted

$ab=cd$ implies $a=xy, b=zt, c=xz, d=yt$ for some integers $x,y,z,t$. Hence $$ a+b+c+d=(x+t)(y+z). $$

share|cite|improve this answer
This may be nitpicking, but $2 \cdot 3 = 3 \cdot 2$ does not imply $2 = xy$. – orlp May 6 '13 at 19:08
@nightcracker: Sorry, I didn't understand. Do you consider the case $a=d=2, b=c=3$? – Boris Novikov May 6 '13 at 19:16
@nightcracker: $a=2$, $b=3$, $c=3$, $d=2$, $x=1$, $y=2$, $z=3$, $t=1$ works. – fgrieu May 6 '13 at 19:59
Of course, pesky $1$ :D Nevermind. – orlp May 6 '13 at 20:07
@nightcracker: Still this answer is more that a bit sketchy, for the construction of $x$, $y$, $z$, $t$ is left as an exercise to the reader. – fgrieu May 6 '13 at 20:53

From $ab=cd$ you have $$(a+b)^2-(a-b)^2=(c+d)^2-(c-d)^2\Rightarrow(a+b)^2-(c+d)^2=(a-b)^2-(c-d)^2$$ Hence we have $$(a+b+c+d)(a+b-c-d)=(a-b+c-d)(a-b-c+d)$$ Now note that $|a+b+c+d|>|a-b+c-d|$ and $|a-b-c+d|$. If $(a+b+c+d)$ was prime then it must divide one of $(a-b+c-d)$ or $(a-b-c+d)$, which is not possible.

share|cite|improve this answer
+1. Beautiful. Though this argument won't work (as it is) for $a,b,c,d \in \mathbb{Z}^+$. – user17762 May 6 '13 at 15:51
It's not hard to show a slightly simpler identity: $a(a+b+c+d)=(a+c)(a+d)$, with the result following from there by a similar size argument. – Thomas Andrews May 6 '13 at 16:01
(+1) Brilliant. – achille hui May 6 '13 at 16:08
@user17762 It's not clear to me what is wrong with the proof in your eyes. – Thomas Andrews May 6 '13 at 16:39
@ThomasAndrews I think he was getting at the fact that one (or both) of $a-b+c-d$ or $a-b-c+d$ may be $0$, in which case $a+b+c+d$ would divide it even if it were prime. Of course, in that case $a,b$ is the same as $c,d$ modulo reordering, so the sum is divisible by 2 and the result still holds. – jerry May 6 '13 at 21:08

From $ab=cd$, We may assume $a=\frac{cd}{b}$. So $M=a+b+c+d = \frac{cd}{b}+b+c+d = \frac{(b+c)(b+d)}{b}$ and so $bM=(b+c)(b+d)$ and $M|(b+c)(b+d)$. We assume that $M$ is not composite, so it is prime. Now we may know that either $b+c$ or $b+d$ is divisible by $M$. So $M\leq b+c$ or $M\leq b+d$ which both result in contradiction because $M=a+b+c+d > b+c$ or $b+d$. So our assumption was wrong and $M$ is a composite number.

share|cite|improve this answer

Hint: Plug $a=\frac{cd}{b}$ into the sum to get


which cannot be prime.

share|cite|improve this answer
Why can you assume that? – Thomas Andrews May 6 '13 at 15:57
It may not be correct if $b+d$ was not divisible by $b$. – CODE May 6 '13 at 16:00
@CODE: See the edit. – Ehsan M. Kermani May 6 '13 at 16:04
Dear @Thomas Andrews: I added the word hint into the body. – Ehsan M. Kermani May 6 '13 at 16:09
@Thomas Andrews: thanks, edited! – Ehsan M. Kermani May 6 '13 at 17:44


$ab$ has to have at least $3$ prime factors.(If $a,b,c,d$ are distinct naturals)

$ab=p_1p_2p_3\dots p_n=cd$

$a=p_1p_2 \dots p_j$

$b=p_{j+1} \dots p_n$

$c=p_kp_{k+1} \dots p_l$

$d=p_1p_2 \dots p_{k-1}p_{l+1}p_{l+2} \dots p_n$

share|cite|improve this answer
Who says they have to be distinct? Take, for example, $a=b=2, c=4, d=1$. – jerry May 6 '13 at 21:29
None says they have to be distinct. You have a more general case if you distinct numbers. – Inceptio May 7 '13 at 3:44

Call the sum $\,f.\ $ Then $\ af = a^2 + \overbrace{ab}^{\Large cd}+ac+ad = \overbrace{(a+c)}^{\large M}\,\overbrace{(a+d)}^{\large N}\,$

By unique factorization $\, a\mid MN\,\Rightarrow\,a = mn,\,\ m\mid M,\, n\mid N,\,$ thus

$\ f = MN/a= (M/m)(N/n).\ $ Each factor is $>1\,$ by $\,m,n \le a < M,N$

share|cite|improve this answer
Imported from a deleted question. – Bill Dubuque Apr 23 '15 at 2:52

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.