Loading [MathJax]/jax/output/SVG/jax.js
Research article

The least common multiple of consecutive terms in a cubic progression

  • Received: 30 October 2019 Accepted: 07 February 2020 Published: 14 February 2020
  • MSC : 11B25, 11N13, 11A05

  • Let k be a positive integer and f(x) a polynomial with integer coefficients. Associated to the least common multiple lcm0ik{f(n+i)}, we define the function Gk,f for all positive integers nNZk,f by Gk,f(n):=ki=0|f(n+i)|lcm0ik{f(n+i)}, where Zk,f:=ki=0{nN:f(n+i)=0}. If f(x)=x, then Farhi showed in 2007 that Gk,f is periodic with k! as its period. Consequently, Hong and Yang improved Farhi's period k! to lcm(1,...,k). Later on, Farhi and Kane confirmed a conjecture of Hong and Yang and determined the smallest period of Gk,f. For the general linear polynomial f(x), Hong and Qian showed in 2011 that Gk,f is periodic and got a formula for its smallest period. In 2015, Hong and Qian characterized the quadratic polynomial f(x) such that Gk,f is almost periodic and also arrived at an explicit formula for the smallest period of Gk,f. If degf(x)3, then one naturally asks the following interesting question: Is the arithmetic function Gk,f almost periodic and, if so, what is the smallest period? In this paper, we asnwer this question for the case f(x)=x3+2. First of all, with the help of Hua's identity, we prove that Gk,x3+2 is periodic. Consequently, we use Hensel's lemma, develop a detailed p-adic analysis to Gk,x3+2 and particularly investigate arithmetic properties of the congruences x3+20(modpe) and x6+1080(modpe), and with more efforts, its smallest period is finally determined. Furthermore, an asymptotic formula for log lcm0ik{(n+i)3+2} is given.

    Citation: Zongbing Lin, Shaofang Hong. The least common multiple of consecutive terms in a cubic progression[J]. AIMS Mathematics, 2020, 5(3): 1757-1778. doi: 10.3934/math.2020119

    Related Papers:

    [1] Guangyan Zhu, Kaimin Cheng, Wei Zhao . Notes on Hong's conjecture on nonsingularity of power LCM matrices. AIMS Mathematics, 2022, 7(6): 10276-10285. doi: 10.3934/math.2022572
    [2] Naif Alotaibi, A. S. Al-Moisheer, Ibrahim Elbatal, Salem A. Alyami, Ahmed M. Gemeay, Ehab M. Almetwally . Bivariate step-stress accelerated life test for a new three-parameter model under progressive censored schemes with application in medical. AIMS Mathematics, 2024, 9(2): 3521-3558. doi: 10.3934/math.2024173
    [3] Kai Wang, Guicang Zhang . Curve construction based on quartic Bernstein-like basis. AIMS Mathematics, 2020, 5(5): 5344-5363. doi: 10.3934/math.2020343
    [4] Yulu Feng, Min Qiu . Some results on p-adic valuations of Stirling numbers of the second kind. AIMS Mathematics, 2020, 5(5): 4168-4196. doi: 10.3934/math.2020267
    [5] Keqiang Li, Shangjiu Wang . Multiple periodic solutions of nonlinear second order differential equations. AIMS Mathematics, 2023, 8(5): 11259-11269. doi: 10.3934/math.2023570
    [6] Hu Jiayuan, Chen Zhuoyu . On distribution properties of cubic residues. AIMS Mathematics, 2020, 5(6): 6051-6060. doi: 10.3934/math.2020388
    [7] Yan Yan . Multiplicity of positive periodic solutions for a discrete impulsive blood cell production model. AIMS Mathematics, 2023, 8(11): 26515-26531. doi: 10.3934/math.20231354
    [8] Shuangnian Hu, Rongquan Feng . The number of solutions of cubic diagonal equations over finite fields. AIMS Mathematics, 2023, 8(3): 6375-6388. doi: 10.3934/math.2023322
    [9] Long Chen, Shaofang Hong . Divisibility among determinants of power matrices associated with integer-valued arithmetic functions. AIMS Mathematics, 2020, 5(3): 1946-1959. doi: 10.3934/math.2020130
    [10] Yonghang Chang, Menglan Liao . Global well-posedness and scattering of the four dimensional cubic focusing nonlinear Schrödinger system. AIMS Mathematics, 2024, 9(9): 25659-25688. doi: 10.3934/math.20241254
  • Let k be a positive integer and f(x) a polynomial with integer coefficients. Associated to the least common multiple lcm0ik{f(n+i)}, we define the function Gk,f for all positive integers nNZk,f by Gk,f(n):=ki=0|f(n+i)|lcm0ik{f(n+i)}, where Zk,f:=ki=0{nN:f(n+i)=0}. If f(x)=x, then Farhi showed in 2007 that Gk,f is periodic with k! as its period. Consequently, Hong and Yang improved Farhi's period k! to lcm(1,...,k). Later on, Farhi and Kane confirmed a conjecture of Hong and Yang and determined the smallest period of Gk,f. For the general linear polynomial f(x), Hong and Qian showed in 2011 that Gk,f is periodic and got a formula for its smallest period. In 2015, Hong and Qian characterized the quadratic polynomial f(x) such that Gk,f is almost periodic and also arrived at an explicit formula for the smallest period of Gk,f. If degf(x)3, then one naturally asks the following interesting question: Is the arithmetic function Gk,f almost periodic and, if so, what is the smallest period? In this paper, we asnwer this question for the case f(x)=x3+2. First of all, with the help of Hua's identity, we prove that Gk,x3+2 is periodic. Consequently, we use Hensel's lemma, develop a detailed p-adic analysis to Gk,x3+2 and particularly investigate arithmetic properties of the congruences x3+20(modpe) and x6+1080(modpe), and with more efforts, its smallest period is finally determined. Furthermore, an asymptotic formula for log lcm0ik{(n+i)3+2} is given.


    The least common multiple of consecutive positive integers was investigated by Chebyshev for the first significant attempt to prove the prime number theorem in [2]. Since then, the least common multiple of any given sequence of positive integers has been an interesting and important topic. Hanson [9] and Nair [20] got the upper and lower bound of lcm1in{i}, respectively. Then Nair's lower bound was extended in [3,4,5,10,11,14,17,22,25]. Goutziers [8] investigated the asymptotic behavior on the least common multiple of a set of integers not exceeding N. Bateman, Kalb and Stenger [1] obtained an asymptotic estimate for the least common multiple of arithmetic progressions. Hong, Qian and Tan [15] got an asymptotic estimate for the least common multiple of the sequence of product of linear polynomials. Qian and Hong [23] studied the asymptotic behavior of the least common multiple of consecutive terms in arithmetic progressions. In [6], Farhi found an interesting relation between the least common multiple and the derivative of integr-valued polynomials. Actually, Farhi [6] showed that lcm1in{i} is the smallest positive integer cn satisfying the property: For any P(x)En, one has cnP(x)En, where En is the set of all the integer-valued polynomials of degree no more than n.

    Let k be a positive integer and f(x) be a polynomial with integer coefficients. Associated to the least common multiple lcm0ik{f(n+i)} of any k+1 consecutive terms in the sequence {f(n)}n=1, we define the arithmetic function Gk,f for all positive integers nNZk,f by

    Gk,f(n):=ki=0|f(n+i)|lcm0ik{f(n+i)},

    where

    Zk,f:=ki=0{nN:f(n+i)=0}.

    If f(x)=x, then Farhi [4] showed that Gk,f is periodic with k! as its period. Consequently, Hong and Yang [16] improved Farhi's period k! to lcm(1,...,k). Later on, Farhi and Kane [7] confirmed the Hong-Yang conjecture in [16] and determined the smallest period of Gk,f. For the general linear polynomial f(x), Hong and Qian [12] showed that Gk,f is periodic and got a formula for its smallest period. In 2012, Qian, Tan and Hong [24] proved that if f(x)=x2+1, then Gk,f is periodic with determining its exact period. In 2015, Hong and Qian [13] characterized the quadratic polynomial f(x) such that Gk,f is almost periodic and also arrived at an explicit formula for the smallest period of Gk,f. One naturally asks the following problem: Let degf(x)3. Is the arithmetic function Gk,f almost periodic and, if so, what is the smallest period? This is a nontrivial and interesting question. In general, it is hard to answer it because few are known about the roots of higher degree congruences modulo prime powers that play key roles in this topic. Among the higher degree sequences, {n3+2}n=1 is the first example one would like to understand. For instance, a renowned conjecture in number theory states that there are infinitely many primes in the cubic sequence {n3+2}n=1.

    In this paper, our main goal is to study the above problem for the cubic sequence {n3+2}n=1. In what follows, for brevity, we write Gk:=Gk,f when f(x)=x3+2. That is, we have

    Gk(n):=ki=0((n+i)3+2)lcm0ik{(n+i)3+2}. (1.1)

    Assume that Gk is periodic and let Pk denote its smallest period. Now we can use Pk to give a formula for lcm0ik{(n+i)3+2} as follows: For any positive integer n, one has

    lcm0ik{(n+i)3+2}=ki=0((n+i)3+2)Gk(nPk),

    where nPk means the least positive integer congruent to n modulo Pk. So it is significant to determine the exact value of the smallest period Pk.

    Before stating the main results of this paper, let us briefly explain the basic new ideas and key techniques in this paper. To deal with the cubic sequence {n3+2}n=1, we have to develop furthermore the techniques presented in previous works [7,12,16,24] and [13] that are far not enough for the current case. In fact, we use Hua's identity [18] to show the periodicity and establish the factorization result to reduce the computation of the smallest period Pk to that of the local periods Pp,k. For the small primes p=2 and 3, we calculate directly Pp,k. For the prime p5, we consider two cases p1(mod6) and p5(mod6). The former case is much more difficult to treat than the latter one. More complicated analysis will be needed in the former case. But for both cases, one needs to study in detail the arithmetic properties of roots of the congruences x3+20(modpe) and x6+1080(modpe), where e1 is an integer and the famous Hensel's lemma is used to lift the roots of the two congruences from modulo p to modulo pe.

    To state our main results, as usual, for any prime number p, we let vp be the normalized p-adic valuation of Q, that is, vp(a)=b if pba. We also let gcd(a,b) denote the greatest common divisor of any integers a and b. For any real number x, by x we denote the largest integer no more than x, we denote by x the smallest integer no less than x. Further, by {x} we denote the fractional part of x, i.e. {x}:=xx. Obviously, 0{x}<1. For any positive integer k, we define the positive integers Rk and Qk by

    Rk:=lcm1ik{i(i6+108)} (1.2)

    and

    Qk:=2(1)k+123{k+13}Rk2v2(Rk)3v3(Rk)p|Rkp1(mod6)2p131(modp)pvp(Rk). (1.3)

    Evidently, vp(Qk)=vp(Rk) for any prime p5(mod6). The main result of this paper can be stated as follows.

    Theorem 1.1. Let k be a positive integer. The arithmetic function Gk is periodic, and its smallest period is equal to R1=109 if k=1, and if k2, then its smallest period equals Qk except that vp(k+1)vp(Qk)1 for at most one prime p5, in which case its smallest period is equal to Qkpvp(Qk).

    Obviously, Theorem 1.1 answers affirmatively the above question for the case where f(x)=x3+2. Moreover, from Theorem 1.1 one can easily deduce the following interesting asymptotic result.

    Theorem 1.2. Let k be any given positive integer. Then the following asymptotic formula holds:

    log lcm0ik{(n+i)3+2}3(k+1)logn  as n.

    This paper is organized as follows. In Section 2, we first use an identity of Hua [18] to show that the arithmetic function Gk is periodic with Rk as a period of it. Then we factor the smallest period Pk of Gk into the product of local period Pp,k with p running over all prime divisors of Rk, and determine the local periods P2,k and P3,k. Later on, with a little more effort, we show that Qk is a period of Gk (see Theorem 2.7 below). Subsequently, we develop in Section 3 a p-adic analysis of the periodic function Gk, and determine the local period Pp,k of Gk for the case p5(mod6). In Section 4, we evaluate the local period Pp,k when p1(mod6) and 2p131(modp). In the process, we need to explore the arithmetic properties of the congruences x3+20(modpe) and x6+1080(modpe). Particularly, we express the smallest positive root of x6+1080(modpe) in the terms of the roots of x3+20(modpe). In the final section, we provide the proofs of Theorems 1.1 and 1.2. Two examples are also given to illustrate the validity of our main result.

    In this section, by using a well-known theorem of Hua in [18], we first prove that Gk is periodic. In the meantime, we also arrive at a nontrivial period of Gk.

    Lemma 2.1. The arithmetic function Gk is periodic, and Rk is a period of Gk.

    Proof. For any positive integer n, it follows from Theorem 7.3 of [18] (see page 11 of [18]) that

    Gk(n)=kt=10i0<...<itk(gcd((n+i0)3+2,...,(n+it)3+2))(1)t1

    and

    Gk(n+Rk)=kt=10i0<...<itk(gcd((n+i0+Rk)3+2,...,(n+it+Rk)3+2))(1)t1.

    Claim that Gk(n+Rk)=Gk(n) from which one can read that Gk is periodic with Rk as its period. It remains to show the claim that will be done in what follows. It is clear that the claim follows if one can prove the following identity holds for all integers i and j with 0i<jk:

    gcd((n+i+Rk)3+2,(n+j+Rk)3+2)=gcd((n+i)3+2,(n+j)3+2). (2.1)

    One can easily check that

    u(n,i,j)((n+i)3+2)u(n,j,i)((n+j)3+2)=(ji)((ji)6+108),

    where u(n,i,j):=4(ij)43(n+2ji)((2n+i+j)(ij)26). So one has

    gcd((n+i)3+2,(n+j)3+2)|(ji)((ji)6+108).

    It then follows from (ji)((ji)6+108)|Rk that

    gcd((n+i)3+2,(n+j)3+2)|Rk. (2.2)

    This implies that

    gcd((n+i)3+2,(n+j)3+2)|(n+i±Rk)3+2

    and

    gcd((n+i)3+2,(n+j)3+2)|(n+j±Rk)3+2.

    One then derives that

    gcd((n+i)3+2,(n+j)3+2)|gcd((n+i+Rk)3+2,(n+j+Rk)3+2) (2.3)

    and

    gcd((n+i)3+2,(n+j)3+2)|gcd((n+iRk)3+2,(n+jRk)3+2). (2.4)

    On the other hand, replacing n by n+Rk in (2.4) gives us that

    gcd((n+i+Rk)3+2,(n+j+Rk)3+2)|gcd((n+i)3+2,(n+j)3+2). (2.5)

    Therefore (2.1) follows immediately from (2.3) and (2.5). Hence the claim is proved.

    This completes the proof of Lemma 2.1.

    For any given prime p, we define the arithmetic function Gp,k for any positive integer n by Gp,k(n):=vp(Gk(n)). Since Gk is a periodic function, Gp,k is periodic for each prime p and Pk is a period of Gp,k. Let Pp,k be the smallest period of Gp,k. Then Pp,k is called p-adic period of Gk. All the p-adic periods Pp,k are called local period of Gk. The following result says that Pp,k is a power of p, and Pk equals the product of all the p-adic periods Pp,k with p being a prime divisor of Rk.

    Lemma 2.2. For any prime p, Pp,k divides pvp(Rk). Furthermore, one has

    Pk=p|RkPp,k.

    Proof. At first, we show that for any prime p, pvp(Rk) is a period of Gp,k. To do so, it is enough to prove that for any given positive integer n and integers i and j with 0i<jk, one has

    vp(gcd((n+i+pvp(Rk))3+2,(n+j+pvp(Rk))3+2))=vp(gcd((n+i)3+2,(n+j)3+2)). (2.6)

    In the following we show that (2.6) is true. Clearly, (2.2) infers that vp(gcd((n+i)3+2,(n+j)3+2))vp(Rk). In other words, we have vp((n+i)3+2)vp(Rk) or vp((n+j)3+2)vp(Rk). From this one can deduce that vp((n+i)3+2)vp((n+i±pvp(Rk))3+2) or vp((n+j)3+2)vp((n+j±pvp(Rk))3+2). It follows that

    vp(gcd((n+i)3+2,(n+j)3+2))=min(vp((n+i)3+2),vp((n+j)3+2)))min{vp((n+i+pvp(Rk))3+2),vp((n+j+pvp(Rk))3+2)}=vp(gcd((n+i+pvp(Rk))3+2,(n+j+pvp(Rk))3+2)) (2.7)

    and

    vp(gcd((n+i)3+2,(n+j)3+2))vp(gcd((n+ipvp(Rk))3+2,(n+jpvp(Rk))3+2)). (2.8)

    Replacing n by n+pvp(Rk) in (2.8) gives us that

    vp(gcd((n+i+pvp(Rk))3+2,(n+j+pvp(Rk))3+2))vp(gcd((n+i)3+2,(n+j)3+2))). (2.9)

    Therefore (2.6) follows immediately from (2.7) and (2.9).

    Now using (2.6) and Hua's identity (Theorem 7.3 of [18]), we can deduce that for any given prime p, Gp,k(n)=Gp,k(n+pvp(Rk)) holds for any positive integer n. Hence pvp(Rk) is a period of Gp,k which implies that Pp,k|pvp(Rk). Therefore Pp,k are pairwise relatively prime for different prime numbers p and Pp,k=1 for those primes pRk. So prime q|RkPq,k|Pk since Pp,k|Pk for each prime p.

    On the other hand, by the definition of Pp,k, we know that for any positive integer n, vp(Gk(n+prime q|RkPq,k))=vp(Gk(n)) holds for all primes p. Hence Gk(n+prime q|RkPq,k)=Gk(n) holds for any positive integer n. That is, p|RkPp,k is a period of Gk, which implies that Pk|prime p|RkPp,k. Hence Pk=p|RkPp,k as desired. Lemma 2.2 is proved.

    In order to determine the smallest period Pk of Gk, by Lemma 2.2 it is enough to determine the exact value of Pp,k for all prime divisors p of Rk. In what follows, we show that Qk is a period of Gk. For arbitrary positive integers n and k, let

    Sk(n):={n3+2,(n+1)3+2,...,(n+k)3+2} (2.10)

    be the set of any k+1 consecutive terms in the cubic progression {m3+2}mN, and

    S(e)k(n):={mSk(n):pe|m}. (2.11)

    Then

    Gp,k(n)=mSk(n)vp(m)maxmSk(n){vp(m)} (2.12)
    =e=1|S(e)k(n)|e=1(1 if pe|m for some mSk(n))=e=1max(0,|S(e)k(n)|1). (2.13)

    We denote lp:=vp(Rk) for any prime p. If there is at most one element divisible by plp+1 in Sk(n) for any positive integer n, then all the terms on the right-hand side of (2.13) are 0 if elp+1. It then follows that for any positive integer n, if there is at most one element divisible by plp+1 in Sk(n), then

    Gp,k(n)=lpe=1fe(n)=lp1e=1fe(n)+flp(n), (2.14)

    where

    fe(n):=max(0,|S(e)k(n)|1). (2.15)

    Clearly, one has that fe(n)=0 if |S(e)k(n)|1, and fe(n)=|S(e)k(n)|1 if |S(e)k(n)|>1.

    Lemma 2.3. We have P2,k=2(1)k+12.

    Proof. Clearly, for any odd integer n, we have v2(n3+2)=0. For any even integer n, letting n=2m gives us that v2(n3+2)=v2((2m)3+2)=v2(8m3+2)=1. So for any positive integer k, one has that maxmSk(n){v2(m)}=1 and

    mSk(n)v2(m)=ki=0v2((n+i)3+2)=v2((n+k)3+2)δk+k12i=0(v2((n+2i)3+2)+v2((n+2i+1)3+2)). (2.16)

    where δk:=1 if 2|k, and 0 otherwise. One can easily see that for any positive integers k and n, exactly one of v2((n+2i)3+2) and v2((n+2i+1)3+2) equals 1 and another one is 0 for all integers i with 0ik12. This implies that v2((n+2i)3+2)+v2((n+2i+1)3+2)=1. Also v2((n+k)3+2)δk=1 happens only when both of k and n are even, and 0 otherwise. Then by (2.12) and (2.16), we derive that

    g2,k(n)=mSk(n)v2(m)maxmSk(n){vp(m)}=k12+v2((n+k)3+2)δk={k12+1,if k and n are even,k12,otherwise={k12,if k is odd,k2,if k and n are even,k21,otherwise.

    Therefore P2,k=1 if 2k, and if k is even, then for any positive integer n, one has g2,k(n+2)=g2,k(n) and g2,k(n+1)g2,k(n). It follows that P2,k=2 if 2|k as required. So Lemma 2.3 is proved.

    Lemma 2.4. We have that P3,1=1 and P3,k=3{k+13} if k2.

    Proof. First of all, we have

    v3(n3+2)={1,if n1(mod3),0,otherwise.

    Since 3 does not divide the difference of (n+1)3+2 and n3+2, at most one of them is divisible by 3. This implies that g3,1(n)=0 for all positive integers n. So P3,1=1 as desired.

    Now let k2. One has that maxmSk(n){v3(m)}=1 and

    mSk(n)v3(m)=ki=0v3((n+i)3+2)=k23i=0(v3((n+3i)3+2)+v3((n+3i+1)3+2)+v3((n+3i+2)3+2))+v3((n+k)3+2)δ(1)k+v3((n+k1)3+2)δ(2)k, (2.17)

    where

    δ(1)k={1,if k0 or 1(mod3),0,otherwise

    and

    δ(2)k={1,if k1(mod3),0,otherwise.

    One can easily see that for any positive integers k and n, exactly one of v3((n+3i)3+2),v3((n+3i+1)3+2) and v3((n+3i+2)3+2) equals 1 and the remaining two are 0 for all integers i with 0ik23. This implies that v3((n+3i)3+2)+v3((n+3i+1)3+2)+v3((n+3i+2)3+2=1. Then by (2.12) applied to p=3 and (2.17), we derive that

    g3,k(n)=k23+v3((n+k)3+2)δ(1)k+v3((n+k1)3+2)δ(2)k. (2.18)

    If k2(mod3), then δ(1)k=δ(2)k=0. So by (2.18), one has g3,k(n)=k23=k23, which infers that P3,k=1 if k2(mod3).

    If k0(mod3), then δ(1)k=1 and δ(2)k=0. If n1(mod3), then v3((n+k)3+2)=1, and 0 otherwise. Again by (2.18), we have

    g3,k(n)=k23+v3((n+k)3+2)={k23+1,if n1(mod3),k23,otherwise={k3,if n1(mod3),k31,otherwise.

    It follows that for any integer t, one has that g3,k(t+3)=g3,k(t) and g3,k(3t)g3,k(3t+1). That is P3,k=3 if k0(mod3).

    If k1(mod3), then δ(1)k=δ(2)k=1. If n0 or 1(mod3), then v3((n+k1)3+2)+v3((n+k)3+2)=1, and 0 otherwise. Then by (2.18), one has

    g3,k(n)=k23+v3((n+k1)3+2)+v3((n+k)3+2)={k23+1,if n0 or 1(mod3),k23,otherwise={k13,if n0 or 1(mod3),k43,otherwise.

    Hence for any integer t, one has that g3,k(t+3)=g3,k(t) and g3,k(3t+1)g3,k(3t+2). Therefore P3,k=3 if k1(mod3).

    Finally, we can conclude that if k2, then

    P3,k={1,if k2(mod3),3,otherwise.

    Namely, P3,k=3{k+13} if k2. This finishes the proof of Lemma 2.4.

    Lemma 2.5. Let p be a prime such that p1(mod6). Then each of the following is true:

    (ⅰ). The congruence x3+20(modp) is solvable if and only if 2p131(modp).

    (ⅱ). The congruence x6+1080(modp) is solvable if and only if 2p131(modp).

    Proof. (ⅰ). Let m and a be positive integers with gcd(a,p)=1. It is well known that xma(modp) has a solution if and only if ap1gcd(p1,m)1(modp). Since p1(mod6), letting m=3 and a=2 gives that x32(modp) has a solution if and only if 2p131(modp). So part (i) is proved.

    (ⅱ). Likewise, x6108(modp) has a solution if and only if (108)p1gcd(6,p1)=(108)p161(modp). It is sufficient to show that (108)p162p13(modp).

    Let (ap) be the Legendre symbol. By the Gauss quadratic reciprocity law, one has

    (p3)=(1)p12312(3p)=(1)p12(3p).

    Then by Euler's criterion, one has

    (3)p12=(1)p123p12(1)p12(3p)=(p3)(13)1(modp)

    since p1(mod6). Therefore (108)p16=2p13(3)p122p13(modp) as desired. This completes the proof of Lemma 2.5.

    Lemma 2.6. If p is a prime number such that p1(mod6) and 2p131(modp), then Pp,k=1.

    Proof. Since 2p131(modp), by Lemma 2.5, x3+20(modp) has no solution. Thus for any positive integer n, one has vp(n3+2)=0. It then follows that Gp,k(n)=vp(Gk(n))=0. So Pp,k=1 as desired. Lemma 2.6 is proved.

    Theorem 2.7. Let k be a positive integer with k2. Then Qk is a period of Gk.

    Proof. By Lemmas 2.2-2.6, one has

    Pk=P2,kP3,k(p|Rkp1(mod6)2p131(modp)Pp,k)(p|Rkp5(mod6)Pp,k)(p|Pkp1(mod6)2p131(modp)Pp,k)=2(1)k+123{k+13}(p|Rkp5(mod6)Pp,k)(p|Rkp1(mod6)2p131(modp)Pp,k). (2.19)

    Since Pp,k is a power of p for each prime p, the product

    (p|Rkp5(mod6)Pp,k)(p|Rkp1(mod6)2p131(modp)Pp,k)

    divides the integer

    Rk2v2(Rk)3v3(Rk)p|Rkp1(mod6)2p131(modp)pvp(Rk).

    Thus Pk|Qk and Qk is a period of Gk. This concludes the proof of Theorem 2.7.

    Consequently, we prove a result which may be of independently interest.

    Lemma 2.8. There is at most one prime p such that vp(k+1)vp(Rk)1.

    Proof. Suppose that there are two distinct primes p and q such that vp(k+1)vp(Rk)1 and vq(k+1)vq(Rk)1. Then k+1p and k+1q. Furthermore, we have

    k+1pvp(Rk)qvq(Rk)max(pq,pvp(Lk)qvq(Lk)),

    where Lk:=lcm1ik{i}.

    If vp(Lk)=0 or vq(Lk)=0, then pk+1 or qk+1. Thus k+1=p or q since k+1p and k+1q. We arrive at a contradiction since k+1pq.

    If vp(Lk)1 and vq(Lk)1, then

    k+1pvp(Lk)qvq(Lk)>min(pvp(Lk)+1,qvq(Lk)+1). (2.20)

    But the inequality pvp(Lk)+1=plogpk+1>plogpk=k together with qvq(Lk)+1=qlogqk+1>qlogqk=k implies that

    min(pvp(Lk)+1,qvq(Lk)+1)>k. (2.21)

    Clearly, (2.20) is contradict to (2.21). So the assumption is not true. Thus Lemma 2.8 is proved.

    In the conclusion of this section, we state two well-known results that are also needed in this paper.

    Lemma 2.9. (Hensel's lemma) (see, for example, [19] or [21]) Let p be a prime and f(x)Z[x] be a polynomial of degree n with leading coefficient not divisible by p. If there exists an integer x1 such that f(x1)0(modp) and f(x1)0(modp), then for every integer k2, there exists an integer xk suck that f(xk)0(modpk) and xkxk1(modpk1).

    Lemma 2.10. (Theorem 7.2 of [18]) Let p be a prime and let a and m be positive integers such that gcd(p,a)=1. Then either the congruence xma(modp) has no solution or it has gcd(m,p1) solutions.

    Throughout this section, we always assume that p is a prime number such p|Rk and p5(mod6).

    Lemma 3.1. Then the congruence x3+20(modp) has exactly one solution in any complete residue system modulo p which is given by x(2p)(2)p+16(modp).

    Proof. By the Euler's criterion, one has (2)p12(2p)(modp). Since -2 is coprime to p, one has (2p)=±1. Then by the Fermat's little theorem, we have

    ((2p)(2)p+16)3=(2p)(2)p+12(2)p12+p+12(2)p(2)(modp).

    So (2p)(2)p+16 is a solution of the congruence x32(modp).

    Since p5(mod6), one has gcd(3,p1)=1. Then by Lemma 2.10, one has that the congruence x32(modp) has only one solution. Hence Lemma 3.1 is proved.

    Lemma 3.2. The congruence x6+1080(modp) has no solution.

    Proof. At first, we have (1p)=(1)p12. Then by the Gauss quadratic reciprocity law, one has

    (108p)=(3p)=(1p)(3p)=(1p)(1)p12(p3)=(p3)=(23)=1

    since p5(mod6). In other words, x2+1080(modp) has no solution. This implies that the congruence x6+1080(modp) has no solution. So Lemma 3.2 is proved.

    Lemma 3.3. Let e and n be any given positive integers. Each of the following holds:

    (ⅰ).There exists exactly one term divisible by pe in any pe consecutive terms of the cubic progression {(n+i)3+2}iN.

    (ⅱ).There is at most one element divisible by pvp(Rk)+1 in Sk(n).

    Proof. (ⅰ). By Lemma 3.1, for any prime p5(mod6), x3+20(modp) has exactly one solution in any given complete residue system modulo p. It then follows immediately from Lemma 2.9 (Hensel's lemma) that for any positive integer e, the congruence x3+20(modpe) has exactly one solution in a complete residue system modulo pe. So there exist exactly one term divisible by pe in any pe consecutive terms of the cubic progression {(n+i)3+2}iN. Part (ⅰ) is proved.

    (ⅱ). Let lp:=vp(Rk). Then by Lemmas 3.1 and 3.2, one has

    lp=vp(lcm1ik{i(i6+108)}=vp(lcm1ik{i})=max1ik{vp(i)} (3.1)

    for all primes p5(mod6). By (3.1), one derives that

    plpk<plp+1 (3.2)

    Suppose that there exist integers i1 and i2 such that 0i1<i2k and (n+i1)3+2(n+i2)3+20(modplp+1). By part (i), one has i2i1plp+1. This is a contradiction since 0i1<i2k<plp+1. Therefore there is at most one term divisible by plp+1 in Sk(n). This finishes the proof of part (ⅱ).

    Lemma 3.4. Let p be a prime number such that p5(mod6) and p|Rk. Then Pp,k=pvp(Rk) except that vp(k+1)vp(Rk)1, in which case one has Pp,k=1.

    Proof. First of all, part (ⅱ) of Lemma 3.3 tells us that there is at most one element divisible by plp+1 in Sk(n) for any positive integer n.

    Now let vp(k+1)vp(Rk)1. The inequality (3.2) implies that plp<k+1plp+1. Thus we can write k+1=vplp for some integer v with 2vp. For any positive integers n and e with 1elp, it follows from Lemma 3.3 (ⅰ) that |S(e)k(n)|=vplpe. Then |S(e)k(n)|=|S(e)k(n+1)|. By (2.15), one has fe(n)=fe(n+1). Thus Pp,k=1 as desired. Lemma 3.4 is true if vp(k+1)vp(Rk)=lp.

    In what follows, we let vp(k+1)<vp(Rk)=lp. Then we can write k+1r(modplp) for some integer r with pvp(k+1)rplp1. So one can write k+1=vplp+r for some integer v. By (3.2), one has 1v<p. Since for all positive integers n and t, (n+tplp1)3+2n3+2(modpe) holds for any integer e with 1elp1, we deduce that |S(e)k(n)|=|S(e)k(n+tplp1)|. It follows that

    lp1e=1fe(n+tplp1)=lp1e=1fe(n). (3.3)

    By Lemma 2.2, Pp,k|plp. Then Pp,k=plp if and only if plp1 is not a period of Gp,k. So it is sufficient to prove that there exist a positive integers n0 such that Gp,k(n0)Gp,k(n0+plp1). By Lemma 3.3, (2.14) and (3.3), this is equivalent to showing that

    flp(n0)flp(n0+plp1). (3.4)

    Clearly, to prove (3.4), it is enough to show that there is a positive integer n such that

    flp(n)>flp(n+plp1). (3.5)

    By Lemma 3.3 (i), the congruence x3+20(modplp) has exactly one solution in a complete residue system modulo plp. Let n1 be a positive integer such that n31+20(modplp). Then any term divisible by plp in the cubic progression {n3+2}nN must be of the form (n1+tplp)3+2 for some integer t. In the following we show that (3.5) is true. For this purpose, we consider the following two cases:

    Case 1. pvp(k+1)rplpplp1. Then n1+k=n1+vplp+r1n1+vplp. It infers that the set of all the terms divisible by plp in the set Sk(n1) contains the set {n31+2,(n1+plp)3+2,,(n1+vplp)3+2}. Therefore flp(n1)=|S(lp)k(n1)|1(v+1)1v.

    On the other hand, since rplpplp1, we have n1+plp1+k=n1+vplp+plp1+r1<n1+(v+1)plp. But n1+plp1>n1. Hence the set of the terms divisible by plp in the set Sk(n1+plp1) is contained in the set {(n1+plp)3+2,(n1+2plp)3+2,,(n1+vplp)3+2}. So |S(lp)k(n1+plp1)|v. It follows that flp(n1+plp1)=max{0,|S(lp)k(n1+plp1)|1}v1. Then picking n=n1 gives us the desired result (3.5). The claim is proved in this case.

    Case 2. plpplp1+1rplp1. Let n2:=n1+plpplp1+1. Then n2n1+plp. Since p is a prime number with p5(mod6), one has plp>2plp1. Then we have

    n2+k=n1+(v+1)plp+rplp1+1>n1+(v+2)plp2plp1>n1+(v+1)plp.

    So the set of all the terms divisible by plp in the set Sk(n2) contains the set {(n1+plp)3+2,,(n1+(v+1)plp)3+2}. Therefore flp(n2)=|S(lp)k(n2)|1(v+1)1v.

    However, one has n2+plp1=n1+plp+1>n1+plp and

    n2+plp1+k=n1+(v+1)plp+rn1+(v+1)plp1<n1+(v+2)plp.

    Then the set of all the terms divisible by plp in the set Sk(n2+plp1) is contained in the set {(n1+2plp)3+2,(n1+3plp)3+2,,(n1+(v+1)plp)3+2}. It implies that |S(lp)k(n2+plp1)|v. Thus flp(n2+plp1)=|S(lp)k(n2+plp1)|1v1. Hence flp(n2)>flp(n2+plp1). So letting n=n2 gives us the desired result (3.5). The claim is still true in this case.

    Finally, by (3.5), we have Pp,kplp1. That is, plp1 is not a period of Gp,k. Thus Pp,k=pvp(Rk) if vp(k+1)<vp(Rk).

    The proof of Lemma 3.4 is complete.

    Throughout this section, we always assume that p is a prime number such that p1(mod6), p|Rk and 2p131(modp). Then gcd(p,108)=1. We begin with the following result.

    Lemma 4.1. Let e and n be positive integers. Each of the following is true:

    (ⅰ). There exist exactly three terms divisible by pe in any pe consecutive terms of the cubic progression {(n+i)3+2}iN.

    (ⅱ). There exist exactly six terms divisible by pe in any pe consecutive terms of the cubic progression {(n+i)6+108}iN.

    Proof. (ⅰ). Since p1(mod6) and 2p131(modp), by Lemmas 2.5 and 2.10, the congruence x3+20(modp) has exactly gcd(3,p1)=3 solutions. It then follows immediately from Lemma 2.9 (Hensel's lemma) that for any positive integer e, the congruence x3+20(modpe) has exactly 3 solutions. Hence there are exactly 3 terms divisible by pe in any pe consecutive terms of the cubic progression {(n+i)3+2}iN. Part (ⅰ) is proved.

    (ⅱ). Likewise, by Lemmas 2.5 to 2.10 we know that the congruence x6+1080(modpe) has exactly gcd(6,p1)=6 solutions. Hence there are exactly 6 terms divisible by pe in any pe consecutive terms of the cubic progression {(n+i)3+2}iN. Part (ⅱ) is proved.

    For any positive integer e, we define dpe to be the smallest positive root of x6+1080(modpe). By Lemma 4.1, one has dpe<pe for any positive integer e. For the simplicity, we write r0:=max1ik{vp(i)}=vp(Lk). Then dpr0<pr0k and vp((dpr0)6+108)r0=max1ik{vp(i)}. Since vp(gcd(i,i6+108))=vp(gcd(i,108))=0 for any positive integer i and vp((dpr0)6+108)max1ik{vp(i6+108)}, we have

    lp=vp(Rk)=max1ik{vp(i(i6+108))}=max1ik{vp(gcd(i,i6+108))+vp(lcm(i,i6+108))}=max1ik{vp(lcm(i,i6+108))}=max1ik{max(vp(i6+108),vp(i))}=max(max1ik{vp(i6+108)},max1ik{vp(i)})=max1ik{vp(i6+108)}. (4.1)

    So j6+1080(modplp) for some integer j with 1jk. By the definition of dplp, one has kjdplp and vp(d6plp+108)lp. Then by the fact that kdplp and (4.1), we have vp(d6plp+108)lp. Therefore lp=vp(d6plp+108).

    Since vp(d6plp+1+108)lp+1 and lp=max1ik{vp(i6+108)}, one deduces that k<dplp+1. Thus we arrive at

    dplpk<dplp+1. (4.2)

    Lemma 4.2. Let p be an odd prime number with 2p131(modp) and e be a positive integer. If x1 and x2 are distinct roots of the congruence x3+20(modpe) in the interval [1,pe], then x21x2+x1x222(modpe) and x2x1 is a root of the congruence x6+1080(modpe).

    Proof. Obviously, one has gcd(x2,p)=gcd(x1,p)=1.

    First, we show that gcd(x2x1,p)=1. Clearly, gcd(x2x1,p)=1 if e=1. If e>1, then by Lemma 2.9 (Hensel's Lemma), there exist two distinct integers x1 and x2 in the interval [1,p] such that x1x1x2x20(modp) and x31+2x32+20(modp). Therefore one has that gcd(x2x1,p)gcd(x2x1,p)0(modp). Thus we arrive at gcd(x2x1,p)=1.

    Subsequently, we prove that

    x21x2+x1x222(modpe). (4.3)

    In fact, by x31+2x32+20(modpe), we deduce that x1(x32x31)=x1(x2x1)(x22+x1x2+x21)(x2x1)(x1x22+x21x22)0(modpe). Since gcd(x2x1,p)=1, we have x1x22+x21x220(modpe). This implies that (4.3) is true.

    Finally, we show that (x2x1)6+1080(modpe). Actually, by (4.3) and noticing that x31x322(modpe), one derives that

    (x2x1)6+108=x626x52x1+15x42x2120x32x31+15x22x416x2x51+x61+1084+12x1x2230x2x218030x1x22+12x2x21+4+10818(2x1x22x21x2)0(modpe).

    In other words, we have x2x1 is a root of the congruence x6+1080(modpe) as desired. This concludes the proof of Lemma 4.2.

    Lemma 4.3. There is at most one element divisible by plp+1 in Sk(n) for any positive integer n.

    Proof. For any positive integer e, by Lemma 4.1, we know that the congruence x3+20(modpe) holds three distinct roots in the interval [1,pe]. We write x1e,x2e and x3e as the three roots of x3+20(modpe) in the interval [1,pe]. Furthermore, by Lemma 2.9, we have xi1xi2xilp(modp) for 1i3. We may let x11<x21<x31. By Lemma 4.2, one has gcd(xi1jxi2j,p)=gcd(xi1j,p)=1 for 1i1i23 and 1jlp.

    Suppose that there exist positive integers n1 and i0 with 1i0k such that

    n31+2(n1+i0)3+20(modplp+1). (4.4)

    By Lemma 2.9 (Hensel's lemma), we know that the terms divisible by plp+1 in the cubic progression {n3+2}nN must be of the form (x1plp+tplp)3+2,(x2plp+tplp)3+2 or (x3plp+tplp)3+2 for some integer t. So there exist j1,j2{1,2,3} and integers t1,t20 such that n1=xj1lp+t1plp and n1+i0=xj2lp+t2plp.

    First, we show that gcd(i0,p)=1. If j1=j2, then i0=(t2t1)plp=tplp where t:=t2t1. By i0k<dplp+1<plp+1, one has tplp<plp+1 which implies that tp1. Then (n1+i0)3+2=n31+3n21tplp+3n1t2p2lp+t3p3lp+23n21tplp0(modplp+1). This is a contradiction with the fact that (n1+i0)3+20(modplp+1). Therefore we must have j1j2. Then by Lemma 4.2, one has gcd(i0,p)=gcd(xj2lpxj1lp,p)=1. Thus n1+i0n1(modplp+1) and so n1+i0 and n1 are distinct roots of the congruence x3+20(modplp+1).

    By Lemma 4.2, one has i0 is a root of the congruence x6+1080(modplp+1), which implies i0dplp+1 It contradicts with the assumption that 1i0k<dplp+1. So the assumption that there exist positive integers n1 and i0 with 1i0k such that (4.4) holds is false. This proves Lemma 4.3.

    Lemma 4.4. Let e be a positive integer and let x1e,x2e and x3e be the three roots of x3+20(modpe) such that x1e<x2e<x3e and x3ex1e<pe. Then dpe=min(x2ex1e,x3ex2e,x1ex3e+pe) and dpepe43.

    Proof. Let y1:=x2ex1e, y2:=x3ex2e, y3:=x3ex1e, y4:=pey1, y5:=pey2 and y6:=pey3. Then 1yi<pe for any integer i with 1i6.

    By Lemma 4.2, one knows that y1,y2 and y3 are the roots of the congruence x6+1080(modpe). We then deduce that y4,y5 and y6 are the roots of x6+1080(modpe) in the interval [1,pe]. So all the yi (1i6) are the roots of the congruence x6+1080(modpe) in the interval [1,pe].

    Now we show that yiyj(modpe) for all integers i and j with 1ij6. It is obvious that y1y3(modpe), y2y3(modpe), y4y6(modpe) and y5y6(modpe). Now we show that y1y2(modpe). If y1y2(modpe), then x1e+x3e2x2e(modpe). By Lemma 4.2, one has x21ex3e+x1ex23e2(modpe). It follows that

    (x1e+x3e)3(2x2e)3=x31e+x33e8x32e+3x21ex3e+3x1ex23e12+3(x21ex3e+x1ex23e)18(modpe).

    In other words, 18|pe, which contradicts with the assumption p1(mod6). So x1e+x3e2x2e(modpe), i.e. y1y2(modpe). This implies that y4y5(modpe).

    Likewise, one has x2e+x3e2x1e(modpe). Then y6y1=pe(x3e+x2e2x1e)0(modpe). Also we have y4y1=pe2(x2ex1e)0(modpe) and y5y1=pe(x3ex1e)0(modpe). That is, y1yi(modpe) for any integer i with 4i6. Similarly, one has y2yi(modpe) and y3yi(modpe) for any integer i with 4i6. Hence yiyj(modpe) for any integers i and j with 1ij6 as one desires. So all the yi (1i6) are pairwise distinct solutions of the congruence x6+1080(modpe) in the interval [1,pe].

    Consequently, we show that dpe=min(y1,y2,y6). In fact, one has y3=y1+y2 and y6=y4y2=y5y1. It follows that max(y1,y2)<y3 and min(y4,y5)>y6. Thus min1i6{yi}=min(y1,y2,y6). So by the definition of dpe, we have

    dpe=min1i6{yi}=min(y1,y2,y6)=min(x2ex1e,x3ex2e,x1ex3e+pe)

    as required.

    Finally, since y1,y2 and y6 are pairwise distinct and y1+y2+y6=pe, we derive that

    pedpe+(dpe+1)+(dpe+2)=3dpe+3.

    It then follows that

    dpe=min(y1,y2,y6)pe33=pe43

    as one desires. This finishes the proof of Lemma 4.4.

    Lemma 4.5. Let k1 be an integer such that vp(k+1)<vp(Rk). Let x1,x2 and x3 be the three positive roots of x3+20(modplp) such that x1<x2<x3 and x3x1plp1. If max(x2x1,x3x2)<x1+plpx3, then there exist positive integers n1 and n2 such that n1n2(modplp1), |S(lp)k(n1)|>|S(lp)k(n2)| and |S(lp)k(n1)|2.

    Proof. By Lemma 4.1 (ⅰ), there exist exactly 3 terms divisible by plp in any plp consecutive terms of the cubic progression {(n+i)3+2}iN. Since x3i+20(modplp) for 1i3 with x1<x2<x3 and x3x1plp1, we have x30+20(modplp) for any integer x0 in the interval (x1,x3) with x0x2. By Lemma 4.1, we know that the terms divisible by plp in the cubic progression {n3+2}nN must be of the form (x1+tplp)3+2, or (x2+tplp)3+2, or (x3+tplp)3+2 with t being an integer. It then follows from Lemma 4.4 that

    dplpplp43<p+23plp1 and dplp+1<p+23plp. (4.5)

    So there exists a positive integer u[1,p+23] such that

    (u1)plp1dplpuplp11. (4.6)

    Since vp(k+1)<lp, one has plp(k+1). Then we can write

    k=vplp+r (4.7)

    for some integers v and r with  0v<p+23 and 0r<plp1. But max(x2x1,x3x2)<x1+plpx3 and (x2x1)+(x3x2)+(x1+plpx3)=plp. Hence

    x1+plpx3plp+33=plp+53. (4.8)

    Further, by Lemma 4.4 and the hypothesis one can write dplp=xi0+1xi0 for i0{1,2}. Let x4:=x1+plp. Then x3<x4. Now we show Lemma 4.5 by considering the following five cases.

    Case 1. 0r<dplp. Then by (4.2) and (4.7), we have r+1dplpvplp+r, and so v1. Let n1:=x3min(plp1,r+1)+1. Then n1x3 and n1+k=x3+vplp+(r+1)min(plp1,r+1)x3+vplp. It follows that the set of all the terms divisible by plp in the set Sk(n1) contains the set {x33+2,(x1+plp)3+2,(x2+plp)3+2,(x3+plp)3+2,,(x1+vplp)3+2,(x2+vplp)3+2,(x3+vplp)3+2}. Thus |S(lp)k(n1)|3v+14 since v1.

    Now picking n2:=n1+plp1 gives us that n2=x3+plp1min(plp1,r+1)+1>x3. But from (4.8) it follows that x1+plpx3>plp1. This together with the assumption x1+plpx3>dplpr+1 implies that x1+plpx3>max(plp1,r+1). Thus x3+max(plp1,r+1)<x1+plp. We then deduce that n2+k=x3+vplp+plp1+(r+1)min(plp1,r+1)=x3+vplp+max(plp1,r+1)<x1+(v+1)plp. Therefore the set of the terms divisible by plp in the set Sk(n2) is contained in the set {(x1+plp)3+2,(x2+plp)3+2,(x3+plp)3+2,,(x1+vplp)3+2,(x2+vplp)3+2,(x3+vplp)3+2}. Thus |S(lp)k(n2)|3v which implies that |S(lp)k(n1)|>|S(lp)k(n2)| as required. So Lemma 4.5 is true if rdplp.

    Case 2. dplpr<plpuplp1. Let n1:=xi0. Since rdplp and dplp=xi0+1xi0, one has n1+k=xi0+r+vplpxi0+dplp+vplp=xi0+1+vplp. It infers that the set of all the terms divisible by plp in the set Sk(n1) contains the set {x3i0+2,x3i0+1+2,x3i0+2+2,,(xi0+(v1)plp)3+2,(xi0+1+(v1)plp)3+2,(xi0+2+(v1)plp)3+2,(xi0+vplp)3+2,(xi0+1+vplp)3+2}. Hence |S(lp)k(n1)|3v+22.

    Now let n2:=xi0+uplp1. By (4.6), one has n2>xi0+dplp=xi0+1. However, it follows from r<plpuplp1 that n2+k=xi0+vplp+uplp1+r<xi0+vplp+uplp1+(plpuplp1)=xi0+(v+1)plp. Thus the set of the terms divisible by plp in the set Sk(n2) is contained in the set {x3i0+2+2,(xi0+plp)3+2,(xi0+1+plp)3+2,(xi0+2+plp)3+2,,(xi0+vplp)3+2,(xi0+1+vplp)3+2,(xi0+2+vplp)3+2}. Thus |S(lp)k(n2)|3v+1. This implies that |S(lp)k(n2)|<|S(lp)k(n1)|. Lemma 4.5 holds in this case.

    Case 3. plpuplp1r<plpdplp1. Let n1:=xi0+1uplp1+1. Since dplp=xi0+1xi0, by (4.6) one has xi0+1xi0uplp11, i.e. n1xi0. But 1up+23 together with p>4 shows that plp>2uplp1. Then it follows from plpuplp1r that n1+k=xi0+1+vplpuplp1+r+1xi0+1+vplpuplp1+(plpuplp1+1)=xi0+1+(v+1)plp2uplp1+1>xi0+1+vplp+1>xi0+1+vplp. This concludes that the set of all the terms divisible by plp in the set Sk(n1) contains the set {x3i0+2,x3i0+1+2,x3i0+2+2,,(xi0+(v1)plp)3+2,(xi0+1+(v1)plp)3+2,(xi0+2+(v1)plp)3+2,(xi0+vplp)3+2,(xi0+1+vplp)3+2}. Thus |S(lp)k(n1)|3v+22.

    On the other hand, let n2:=xi0+1+1. Then n2>xi0+1. Since r<plpdplp1=plp(xi0+1xi0)1, one has n2+k=xi0+1+vplp+r+1<xi0+1+vplp+(plp(xi0+1xi0))=xi0+(v+1)plp. Hence the set of the terms divisible by plp in the set Sk(n2) is contained in the set {x3i0+2+2,(xi0+plp)3+2,(xi0+1+plp)3+2,(xi0+2+plp)3+2,,(xi0+vplp)3+2,(xi0+1+vplp)3+2,(xi0+2+vplp)3+2}. So |S(lp)k(n2)|3v+1. Thus Lemma 4.5 is proved in this case.

    Case 4. plpdplp1r<plpplp1. Let n1:=xi0. We assert that xi0+2xi0<plpdplp. In fact, if i0=1, then by Lemma 4.4, one has dplp=min(x2x1,x3x2) since max(x2x1,x3x2)<x1+plpx3. So dplp<x1+plpx3 and the assertion follows immediately. If i0=2, then dplp=x3x2x2x1. But the proof of Lemma 4.4 tells us that x3x2x2x1(modplp). Thus x3x2<x2x1 that implies that dplp<x2x1. It then follows from x4=x1+plp that x4x2<plpdplp. The claim is proved.

    Now by the claim and the hypothesis that plpdplp1r, we derive that xi0+2xi0+1plpdplpr+1. Then n1+k=xi0+vplp+rxi0+vplp+(xi0+2xi0)=xi0+2+vplp. It implies that the set of all the terms divisible by plp in the set Sk(n1) contains the set {x3i0+2,x3i0+1+2,x3i0+2+2,(xi0+plp)3+2,,(xi0+vplp)3+2,(xi0+1+vplp)3+2,(xi0+2+vplp)3+2}. Thus |S(lp)k(n1)|3v+3.

    Now let n2:=xi0+plp1. Then n2>xi0. Since r<plpplp1, one has n2+k=xi0+vplp+plp1+r<xi0+vplp+plp1+(plpplp1)=xi0+(v+1)plp. Hence the set of the terms divisible by plp in the set Sk(n2) is contained in the set {x3i0+1+2,x3i0+2+2,(xi0+plp)3+2,(xi0+1+plp)3+2,(xi0+2+plp)3+2,,(xi0+vplp)3+2,(xi0+1+vplp)3+2,(xi0+2+vplp)3+2}. Therefore |S(lp)k(n2)|3v+2 that infers that the desired result |S(lp)k(n2)|<|S(lp)k(n1)| is true. Hence Lemma 4.5 is proved in this case.

    Case 5. plpplp1r<plp1. Let n1:=x3+plp1+1. By (4.8), one has x1+plpx3>plp1+1. Then x1+plpn1=(x1+plpx3)(plp1+1)>0, that is, n1<x1+plp. Moreover, the assumption plpplp1r implies that n1+k=x3+plp1+vplp+r+1>x3+plp1+vplp+(plpplp1)=x3+(v+1)plp. This infers that the set of all the terms divisible by plp in the set Sk(n1) contains the set {(x1+plp)3+2,(x2+plp)3+2,(x3+plp)3+2,,(x1+(v+1)plp)3+2,(x2+(v+1)plp)3+2,(x3+(v+1)plp)3+2}. Hence |S(lp)k(n1)|3v+3.

    Let n2:=x3+1. Then n2>x3. Since r<plp1, we deduce that n2+k=x3+vplp+r+1<x3+vplp+(plp1)+1=x3+(v+1)plp. Then the set of the terms divisible by plp in the set Sk(n2) is contained in the set {(x1+plp)3+2,(x2+plp)3+2,(x3+plp)3+2,,(x1+(v+1)plp)3+2,(x2+(v+1)plp)3+2}. Thus |S(lp)k(n2)|3v+2. So |S(lp)k(n2)|<|S(lp)k(n1)| as required.

    This completes the proof of Lemma 4.5.

    Lemma 4.6. Let p be a prime number such that p1(mod6), p|Rk and 2p131(modp). Then Pp,k=pvp(Rk) except that vp(k+1)vp(Rk)1, in which case one has Pp,k=1.

    Proof. First of all, we let vp(k+1)vp(Rk):=lp. Then plp|(k+1). But k+1plp+1. So we can write k+1=vplp for some positive integer v with 1vp. For all positive integers n and e with 1elp, by Lemma 4.1 one deduces |S(e)k(n)|=3vplpe. Thus |S(e)k(n)|=|S(e)k(n+1)|. By (2.15), we deduce that fe(n)=fe(n+1). That is, Pp,k=1 if vp(k+1)vp(Rk). So Lemma 4.6 is true if vp(k+1)vp(Rk).

    In what follows, we let vp(k+1)<lp. By Lemma 2.2, plp is a period of Gp,k. So Pp,k=plp if and only if plp1 is not a period of Gp,k. In the following, we show that plp1 is not a period of Gp,k.

    Let x1,x2 and x3 be the three positive roots of x3+20(modplp) such that x1<x2<x3 and x3x1plp1. Then x1+plp and x2+plp are the roots of x3+20(modplp) and x2<x3<x1+plp<x2+plp. By the proof of Lemma 4.4, one knows that x2x1,x3x2 and x1+plpx3 are distinct roots of the congruence x6+1080(modplp). Then exactly one of the following three cases happens:

    (ⅰ). max(x2x1,x3x2)<x1+plpx3,

    (ⅱ). max(x3x2,x1+plpx3)<x2x1, which is equivalent to max(x3x2,(x1+plp)x3)<x2+plp(x1+plp),

    (ⅲ). max(x2x1,x1+plpx3)<x3x2, which is equivalent to max((x1+plp)x3,(x2+plp)(x1+plp))<x3+plp(x2+plp).

    So with Lemma 4.5 applied to the three roots x1,x2 and x3 of x3+20(modplp) for case (ⅰ), and with Lemma 4.5 applied to the three roots x2,x3 and x1+plp of x3+20(modplp) for case (ⅱ), and with Lemma 4.5 applied to the three roots x3,x1+plp and x2+plp of x3+20(modplp) for case (ⅲ), Lemma 4.5 tells us that there exist positive integers n1 and n2 such that n1n2(modplp1), |S(lp)k(n1)|>|S(lp)k(n2)| and |S(lp)k(n1)|2. Then one can deduce that

    flp(n1)>flp(n2). (4.9)

    On the other hand, one has n31+2n32+2(modpe) for any integer e with 1elp1. Hence |S(e)k(n1)|=|S(e)k(n2)|. It then follows that

    lp1e=1fe(n1)=lp1e=1fe(n2). (4.10)

    By Lemma 4.3, |S(lp+1)k(n)|1 for any positive integer n. Then we can use (2.14) and put (4.9) and (4.10) together to obtain that

    Gp,k(n1)=lpe=1fe(n1)>lpe=1fe(n2)=Gp,k(n2).

    Thus plp1 is not a period of Gp,k. It concludes that Pp,k=plp if vp(k+1)<vp(Rk)=lp.

    So Lemma 4.6 is proved.

    Using the lemmas presented in previous sections, we are now in a position to prove Theorems 1.1 and 1.2. We begin with the proof of Theorem 1.1.

    Proof of Theorem 1.1. First of all, by Lemma 2.1, we know that Gk is periodic.

    Consequently, since R1=109 is a prime number with 1091(mod6),

    2109131(mod109)

    and

    0=v109(2)<v109(109)=1,

    one derives from Lemma 4.6 that P109,1=109. It then follows from Lemma 2.2 that P1=109.

    Now let k2. By Lemmas 2.3 and 2.4, one has

    P2,k=2(1)k+12

    and

    P3,k=3{k+13}.

    Further, if p1(mod6) and 2p131(modp), Lemma 2.6 tells us that Pp,k=1.

    By Lemma 2.8, we know that there is at most one prime p05 such that vp0(k+1)vp0(Rk)1. If p05(mod6), then by Lemma 3.4 one has Pp0,k=1. If p01(mod6), then we can deduce from Lemmas 2.6 and 4.6 that Pp0,k=1. For all other primes q5 with q|Rk, one derives from Lemmas 3.4 and 4.6 that Pq,k=qvq(Rk).

    Finally, by Lemma 2.2 one then derives that the smallest period Pk of Gk is equal to

    Qk:=2(1)k+123{k+13}Rk2v2(Rk)3v3(Rk)p|Rkp1(mod6)2p131(modp)pvp(Rk)

    except that vp(k+1)vp(Qk)1 for at most one prime p5, in which case its smallest period Pk equals Qkpvp(Qk).

    This finishes the proof of Theorem 1.1.

    Consequently, we give the proof of Theorem 1.2.

    Proof of Theorem 1.2. At first, Gk is periodic by Theorem 1.1. Then for any positive integer n, one has Gk(n)M:=max1mPk{Gk(m)}. So one deduces that

    log(ki=0((n+i)3+2))logMlog lcm0ik{(n+i)3+2}log(ki=0((n+i)3+2)).

    However, one has

    log(ki=0((n+i)3+2))logM=3(k+1)logn+ki=0log(1+3in+3i2n2+i3+2n3)logM.

    It implies that

    limnlog(ki=0((n+i)3+2))logM3(k+1)logn=1.

    On the other hand, we have

    limnlog(ki=0((n+i)3+2))3(k+1)logn=limn(1+ki=0log(1+3in+3i2n2+i3+2n3)3(k+1)logn)=1.

    It then follows that

    limnlog lcm0ik{(n+i)3+2}3(k+1)logn=1

    as one desires. The proof of Theorem 1.2 is complete.

    By Theorem 1.1, we can easily find infinitely many positive integers k such that Pk=Qk as the following example shows.

    Example 5.1. If k+1 has no prime factors congruent to 5 modulo 6, and k+1 has no prime factors p such that p1(mod6) with 2p131(modp), then Pk=Qk by Theorem 1.1. For instance, if k+1 equals 6r with r being a positive integer, then Pk=Qk.

    On the other hand, there are also infinitely many positive integers k such that Pk equals Qk divided by a power of one prime p. Moreover, by Theorem 1.1 we can present the following proposition that gives us such an example.

    Proposition 5.2. If k+1 is equal to tpe for any positive integer e, where p is a prime number with p5(mod6) and t{2,,p1}, then Pk=Qk/pe.

    Remark 5.3. Let k be a positive integer and f(x) be a polynomial with integer coefficients. Let the arithmetic function Gk,f be defined as in the Introduction section. If f(x) is linear, then it was proved in 2011 by Hong and Qian [12] that Gk,f is periodic with determination of its smallest period. In 2015, Hong and Qian [13] characterized the quadratic polynomial f(x) such that Gk,f is almost periodic and an explicit and complicated formula for the smallest period of Gk,f is obtained too. Now let degf(x)3. If f(x)=x3+2, then by Theorem 1.1 of this paper, we know that Gk,f is a periodic function. Furthermore, Theorem 1.1 gives us an explicit formula for its smallest period. By developing the methods presented in [12], [13] and [24] and in this paper, we can show that the arithmetic function Gk,f is periodic when f(x) is irreducible. One can also find reducible polynomials f1(x) and f2(x) such that Gk,f1 is almost periodic and Gk,f2 is not almost periodic. For which reducible polynomials f(x), the arithmetic function Gk,f is almost periodic? What is the smallest period of Gk,f if Gk,f is almost periodic? They were answered in [12], [13] and in this paper when degf(x){1,2} and f(x)=x3+2. However, these problems are kept widely open when f(x)x3+2 and degf(x)3.

    The authors would like to thank the anonymous referees for careful reading of the manuscript and helpful comments and corrections that improved the presentation of this paper. Hong was supported partially by National Science Foundation of China Grant # 11771304. Lin was supported partially by Research Fund of Panzhihua University Grant #2015YB41 and Doctoral Research Initiation Fund Project of Panzhihua University.

    We declare that we have no conflict of interest.



    [1] P. Bateman, J. Kalb and A. Stenger, A limit involving least common multiples, Amer. Math. Monthly, 109 (2002), 393-394.
    [2] P. L. Chebyshev, Memoire sur les nombres premiers, J. Math. Pures Appl., 17 (1852), 366-390.
    [3] B. Farhi, Minorations non triviales du plus petit commun multiple de certaines suites finies d'entiers, C. R. Acad. Sci. Paris, Ser. I, 341 (2005), 469-474. doi: 10.1016/j.crma.2005.09.019
    [4] B. Farhi, Nontrivial lower bounds for the least common multiple of some finite sequences of integers, J. Number Theory, 125 (2007), 393-411. doi: 10.1016/j.jnt.2006.10.017
    [5] B. Farhi, An identity involving the least common multiple of binomial coeffcients and its application, Amer. Math. Monthly, 116 (2009), 836-839. doi: 10.4169/000298909X474909
    [6] B. Farhi, On the derivatives of the integer-valued polynomials, arXiv:1810.07560.
    [7] B. Farhi and D. Kane, New results on the least common multiple of consecutive integers, Proc. Amer. Math. Soc., 137 (2009), 1933-1939.
    [8] C. J. Goutziers, On the least common multiple of a set of integers not exceeding N, Indag. Math., 42 (1980), 163-169.
    [9] D. Hanson, On the product of the primes, Canad. Math. Bull., 15 (1972), 33-37. doi: 10.4153/CMB-1972-007-7
    [10] S. F. Hong and W. D. Feng, Lower bounds for the least common multiple of finite arithmetic progressions, C. R. Acad. Sci. Paris, Ser. I, 343 (2006), 695-698. doi: 10.1016/j.crma.2006.11.002
    [11] S. F. Hong, Y. Y. Luo, G. Y. Qian, et al. Uniform lower bound for the least common multiple of a polynomial sequence, C.R. Acad. Sci. Paris, Ser. I, 351 (2013), 781-785. doi: 10.1016/j.crma.2013.10.005
    [12] S. F. Hong and G. Y. Qian, The least common multiple of consecutive arithmetic progression terms, Proc. Edinb. Math. Soc., 54 (2011), 431-441. doi: 10.1017/S0013091509000431
    [13] S. F. Hong and G. Y. Qian, The least common multiple of consecutive quadratic progression terms, Forum Math., 27 (2015), 3335-3396.
    [14] S. F. Hong and G. Y. Qian, New lower bounds for the least common multiple of polynomial sequences, J. Number Theory, 175 (2017), 191-199. doi: 10.1016/j.jnt.2016.11.026
    [15] S. F. Hong, G. Y. Qian and Q. R. Tan, The least common multiple of a sequence of products of linear polynomials, Acta Math. Hungar., 135 (2012), 160-167. doi: 10.1007/s10474-011-0173-4
    [16] S. F. Hong and Y. J. Yang, On the periodicity of an arithmetical function, C. R. Acad. Sci. Paris Sér. I, 346 (2008), 717-721.
    [17] S. F. Hong and Y. J. Yang, Improvements of lower bounds for the least common multiple of arithmetic progressions, Proc. Amer. Math. Soc., 136 (2008), 4111-4114. doi: 10.1090/S0002-9939-08-09565-8
    [18] L.-K. Hua, Introduction to number theory, Springer-Verlag, Berlin Heidelberg, 1982.
    [19] N. Koblitz, p-Adic numbers, p-adic analysis, and zeta-functions, Springer-Verlag, Heidelberg, 1977.
    [20] M. Nair, On Chebyshev-type inequalities for primes, Amer. Math. Monthly, 89 (1982), 126-129. doi: 10.1080/00029890.1982.11995398
    [21] J. Neukirch, Algebraic number theory, Springer-Verlag, 1999.
    [22] S. M. Oon, Note on the lower bound of least common multiple, Abstr. Appl. Anal., 2013.
    [23] G. Y. Qian and S. F. Hong, Asymptotic behavior of the least common multiple of consecutive arithmetic progression terms, Arch. Math., 100 (2013), 337-345. doi: 10.1007/s00013-013-0510-7
    [24] G. Y. Qian, Q. R. Tan and S. F. Hong, The least common multiple of consecutive terms in a quadratic progression, Bull. Aust. Math. Soc., 86 (2012), 389-404. doi: 10.1017/S0004972712000202
    [25] R. J. Wu, Q. R. Tan and S. F. Hong, New lower bounds for the least common multiple of arithmetic progressions, Chinese Annals of Mathematics, Series B, 34 (2013), 861-864. doi: 10.1007/s11401-013-0805-9
  • Reader Comments
  • © 2020 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(4274) PDF downloads(366) Cited by(0)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog