Can any even number be written in the form of $(n-1)A+B$, where $A$ and $B$ are primes and $n$ is an even number? $$ C= (n-1)A+B $$ Here $C$ is a even number and $A$ and $B$ are prime and $n$ is any even number such that $n < \sqrt{C}$ (or you can say that $n$ is the number of primes used in the eqn)

Is this true? Because if Goldbach's conjecture is true, I presume this should also be true, right?

  • $\begingroup$ what about C=6, n=4? $\endgroup$ – wxu Dec 8 '12 at 16:38
  • $\begingroup$ @wxu: I think the OP is asking if this can be done for every even number, not just for some (positive) even number. $\endgroup$ – hardmath Dec 8 '12 at 16:43
  • $\begingroup$ @wxu May be I can modify that stating n should be less than the squareroot of c [n < sqrt(c)] $\endgroup$ – Shan Dec 8 '12 at 17:02
  • $\begingroup$ What about C=66, n=4 for your modifying case? $\endgroup$ – wxu Dec 8 '12 at 18:46
  • $\begingroup$ @Wxu n can be any even number than is less than 8 so it can be also 6 and need not be 4 and it can be also 2 so that 66=(6-1)11+11 $\endgroup$ – Shan Dec 9 '12 at 4:59

Chen's theorem, that every sufficiently large even number is the sum of a prime and another number with at most two prime factors, is a result in this direction. That is, taking $n=2$ in your expression accounts for those even numbers which are a sum of two (odd) primes, and taking $n=p+1$ where an even number is a sum of a prime and a product $pq$ of two odd primes (semiprime or "almost prime") accounts for the others.

Some work has been done to improve on Chen's work, but I don't know of an explicit bound beyond which his sieve methods show this representation holds.

Edit: The construction above almost meets the bound $n^2 < C = 2m$ if we choose $p$ to be the lesser prime factor of the semiprime summand (if applicable). Unless the factors are equal (the semiprime is a square), $(p+1)^2 < C$.

Without the restriction on $n$ I don't think your result is difficult to prove for $C \gt 4$. Consider the so-called Bertrand's postulate that a prime exists between $m$ and even number $2m$. Let $q$ be the least such prime, so that $2m-q$ is an odd number. If this is prime, we are done (as explained above for $n=2$ in your notation). If this is an odd number greater than $3$, then is has an odd prime factor $p$, and again we are done as before ($2m = ((2m-q)/p)*p + q$ and the factor $(2m-q)/p$ is (being odd) one less than an even $n$).

The only difficulty I can see is with very small even numbers. That is, we cannot express $2$ or $4$ as sums of two (positive) odd primes, nor can we express them as a sum of the kind you ask about (if we restrict ourselves to positive summands).

  • $\begingroup$ I see that some restrictions on the size of $n$ have been added to the Question since I first posted, so my "answer" may not apply to the revised version. Certainly restricting $n$ makes the Question more interesting. $\endgroup$ – hardmath Dec 8 '12 at 17:17
  • $\begingroup$ Wxu's comment made me think more about it $\endgroup$ – Shan Dec 8 '12 at 17:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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