Suppose that N is a sufficiently large real number. In this paper it is proved that for 2<c<990479, the Diophantine equation
[pc1]+[pc2]+[pc3]+[pc4]+[pc5]=N
is solvable in primes p1,p2,p3,p4,p5 such that each of the numbers pi+2,i=1,2,3,4,5 has at most [62273960−1916c] prime factors.
Citation: Liuying Wu. On a Diophantine equation involving fractional powers with primes of special types[J]. AIMS Mathematics, 2024, 9(6): 16486-16505. doi: 10.3934/math.2024799
[1] | Xinyan Li, Wenxu Ge . A Diophantine approximation problem with unlike powers of primes. AIMS Mathematics, 2025, 10(1): 736-753. doi: 10.3934/math.2025034 |
[2] | Jing Huang, Qian Wang, Rui Zhang . On a binary Diophantine inequality involving prime numbers. AIMS Mathematics, 2024, 9(4): 8371-8385. doi: 10.3934/math.2024407 |
[3] | Jing Huang, Ao Han, Huafeng Liu . On a Diophantine equation with prime variables. AIMS Mathematics, 2021, 6(9): 9602-9618. doi: 10.3934/math.2021559 |
[4] | Wenpeng Zhang, Jiafan Zhang . The hybrid power mean of some special character sums of polynomials and two-term exponential sums modulo $ p $. AIMS Mathematics, 2021, 6(10): 10989-11004. doi: 10.3934/math.2021638 |
[5] | Bingzhou Chen, Jiagui Luo . On the Diophantine equations $x^2-Dy^2=-1$ and $x^2-Dy^2=4$. AIMS Mathematics, 2019, 4(4): 1170-1180. doi: 10.3934/math.2019.4.1170 |
[6] | Li Zhu . On pairs of equations with unequal powers of primes and powers of 2. AIMS Mathematics, 2025, 10(2): 4153-4172. doi: 10.3934/math.2025193 |
[7] | Wenpeng Zhang, Yuanyuan Meng . On the sixth power mean of one kind two-term exponential sums weighted by Legendre's symbol modulo $ p $. AIMS Mathematics, 2021, 6(7): 6961-6974. doi: 10.3934/math.2021408 |
[8] | Jinyan He, Jiagui Luo, Shuanglin Fei . On the exponential Diophantine equation $ (a(a-l)m^{2}+1)^{x}+(alm^{2}-1)^{y} = (am)^{z} $. AIMS Mathematics, 2022, 7(4): 7187-7198. doi: 10.3934/math.2022401 |
[9] | Cheng Feng, Jiagui Luo . On the exponential Diophantine equation $ \left(\frac{q^{2l}-p^{2k}}{2}n\right)^x+(p^kq^ln)^y = \left(\frac{q^{2l}+p^{2k}}{2}n\right)^z $. AIMS Mathematics, 2022, 7(5): 8609-8621. doi: 10.3934/math.2022481 |
[10] | Shujie Zhou, Li Chen . On the sixth power mean values of a generalized two-term exponential sums. AIMS Mathematics, 2023, 8(11): 28105-28119. doi: 10.3934/math.20231438 |
Suppose that N is a sufficiently large real number. In this paper it is proved that for 2<c<990479, the Diophantine equation
[pc1]+[pc2]+[pc3]+[pc4]+[pc5]=N
is solvable in primes p1,p2,p3,p4,p5 such that each of the numbers pi+2,i=1,2,3,4,5 has at most [62273960−1916c] prime factors.
For a fixed integer k≥1 and sufficiently large integer N, the well-known Waring-Goldbach problem is devoted to investigating the solvability of the following Diophantine equality
N=pk1+pk2+⋯+pks | (1.1) |
in prime variables p1,p2,…,ps. Numerous mathematicians have derived many splendid results in this field. For instance, in 1937, Vinogradov [25] proved that such a representation of the type (1.1) exists for every sufficiently large odd integer N with k=1,s=3. Later in 1938, based upon Vinogradov's work, Hua [9] showed that (1.1) is solvable for every sufficiently large integer N satisfying that N≡5(mod24) with k=2,s=5.
In 1933, Segal [21,22] studied the following anolog of the well-known Waring problem. Suppose that c>1 and c∉N; there exists a positive integer s=s(c) such that for every sufficiently large natural number N, the equation
N=[mc1]+[mc2]+⋯+[mcs] |
has a solution with m1,m2,…,ms integers, where [t] denotes the integral part of any t∈R.
To obtain a result that is analogous to the ternary Goldbach problem, in 1995, Laporta and Tolev [13] considered the equation
[pc1]+[pc2]+[pc3]=N, | (1.2) |
where p1,p2,p3 are prime numbers, c∈R,c>1,N∈N, and [t] denotes the integral part of t. They proved that if 1<c<1716 and N is a sufficiently large integer, then the Eq (1.2) has a solution in prime numbers p1,p2,p3. Later, the upper bound of c was enlarged to
1211,258235,137119,31132703,35813106 |
by Kumchev and Nedeva [12], Zhai and Cao [27], Cai [6], Li and Zhang [15], and Baker [2], successively and respectively.
On the other hand, as an analogue of Hua's theorem on five prime squares, Li and Zhang [14] first studied the solvability of the Diophantine equation
[pc1]+[pc2]+[pc3]+[pc4]+[pc5]=N | (1.3) |
in prime numbers p1,p2,p3,p4,p5. They proved that if 1<c<41090541999527,c≠2 and N is a sufficiently large integer, then the Eq (1.3) has a solution in prime numbers p1,p2,p3,p4,p5. Later this result was improved by Li [17] who enlarged the upper bound for c to 408197, and by Baker [2] who replaced 408197 by 609293.
For any natural number r, let Pr denote an almost-prime with at most r prime factors, counted according to multiplicity. There are many papers that are devoted to the study of problems involving primes of a special type. In 1973, Chen [4] established that there exist infinitely many primes p such that p+2 has at most 2 prime factors. In 2000, Tolev [24] proved that for every sufficiently large integer N≡3(mod6), the equation
p1+p2+p3=N | (1.4) |
has a solution in prime numbers p1,p2,p3 such that p1+2∈P2,p2+2∈P5,p3+2∈P7. After that, this result was improved by some mathematicians, and the best result in this field was obtained by Matomäki and Shao [18], who showed that for every sufficiently large integer N≡3(mod6) the Eq (1.4) has a solution in prime numbers p1,p2,p3 such that p1+2,p2+2,p3+2∈P2.
Bearing in mind the result of [18], it is natural for us to conjecture that if c is close to 1, then the Eq (1.2) is solvable in primes p1,p2,p3 such that pi+2∈P2. An attempt to establish this kind of the result was first made by Petrov [19], who showed that, for 1<c<1716 and every sufficiently large integer N, the Eq (1.2) is solvable in prime numbers p1,p2,p3 such that each of the numbers pi+2 has at most [9517−16c] prime factors, counted according to multiplicity. Recently, Li et al. [16] improved Petrov's result; they extended the range of c to 1<c<21731930 and reduced the number of prime factors of pi+2,i=1,2,3 to [113874346−3860c].
Referencing Hua's work, Tolev [24] also showed that for every sufficiently large integer N≡5(mod24), the equation
p21+p22+p23+p24+p25=N | (1.5) |
has a solution in prime numbers p1,p2,p3,p4,p5 such that p1+2∈P2,p2+2∈P2,p3+2∈P5,p4+2∈P5 and p5+2∈P8. And later in 2009, Cai and Lu [5] improved Tolev's result by showing that the Eq (1.5) has a solution in prime numbers p1,p2,p3,p4,p5 such that p1+2∈P2,p2+2∈P2,p3+2∈P4,p4+2∈P4 and p5+2∈P5. Motivated by Petrov [19] and Tolev [24], it is reasonable to conjecture that if N is a sufficiently large natural number and c is close to 2, then the Eq (1.3) has a solution in prime numbers p1,p2,p3,p4,p5 such that pi+2 are almost-primes of a certain fixed order.
In this paper, we shall prove the following result.
Theorem 1.1. Suppose that 2<c<990479 and let N be a sufficiently large natural number. Then the equation (1.3) has a solution in prime numbers p1,p2,p3,p4,p5 such that each of the numbers p1+2,p2+2,p3+2,p4+2 and p5+2 has at most [62273960−1916c] prime factors, counted with the multiplicity.
Throughout this paper, the letter p, with or without subscript, always stand for prime numbers. We use ε to denote a sufficiently small positive number, and the value of ε may change from statement to statement. As usual, we use μ(n),Λ(n),φ(n) and τ(n) to denote Möbius' function, von Mangolds' function, Euler's function and the Dirichlet divisor function, respectively. We write f=O(g) or, equivalently, f≪g if |f|≤Cg for some positive number C. If we have simultaneously, that A≪B and B≪A, then we shall write A≍B. Moreover, we shall use (m,n) and [m,n] for the greatest common divisor and the least common multiple of the integers m and n, respectively. And we use e(α) to denote e2πiα. In addition, we define
2<c<990479,X=(N3)1c,δ=990479−c,ξ=3c2−52,η=4δ13,D=Xδ,z=Xη,τ=Xξ−c,P(z)=∏2<p<zp,logp=5∏j=1(logpj),λ±(d)Rosser's weights of orderD. | (2.1) |
Lemma 2.1. Suppose that D>4 is a real number and let λ±(d) represent the Rosser functions of level D. Then we have the following properties:
(1) For any positive integer d we have
|λ±(d)|≤1,λ±(d)=0ifd>Dorμ(d)=0. |
(2) If n∈N then
∑d∣nλ−(d)≤∑d∣nμ(d)≤∑d∣nλ+(d). | (2.2) |
(3) If z∈R and if
P(z)=∏2<p<zp,B=∏2<p<z(1−1p),N±=∑d∣P(z)λ±(d)φ(d),s0=logDlogz, | (2.3) |
then we have
B≤N+≤B(F(s0)+O((logD)−1/3)), |
B≥N−≥B(f(s0)+O((logD)−1/3)), |
where F(s) and f(s) denote the classical functions in the linear sieve theory that are respectively defined by
F(s)=2eγs(1+∫s−12log(t−1)tdt),3<s≤5 |
and
f(s)=2eγlog(s−1)s,2<s≤4. |
Here γ denotes the Euler constant.
Proof. This is a special case of the work by Greaves [8].
Lemma 2.2. Let
Λi=∑d|(pi+2,P(z))μ(d),Λ±i=∑d|(pi+2,P(z))λ±(d),i=1,2,3,4,5. |
Then we have
Λ1Λ2Λ3Λ4Λ5≥Λ−1Λ+2Λ+3Λ+4Λ+5+Λ+1Λ−2Λ+3Λ+4Λ+5+Λ+1Λ+2Λ−3Λ+4Λ+5+Λ+1Λ+2Λ+3Λ−4Λ+5+Λ+1Λ+2Λ+3Λ+4Λ−5−4Λ+1Λ+2Λ+3Λ+4Λ+5. |
Proof. The proof is the same as in Lemma 13 of [3].
Lemma 2.3. Suppose that f(x):[a,b]→R has continuous derivatives of arbitrary order on [a,b], where 1≤a<b≤2a. Suppose further that
|f(j)(x)|≍λ1a1−j,j≥1,x∈[a,b]. |
Then for any exponential pair (κ,λ), we have
∑a<n≤be(f(n))≪λκ1aλ+λ−11. |
Proof. See (3.3.4) of [7].
Lemma 2.4. For any complex number zn, we have
|∑a<n≤bzn|2≤(1+b−aQ)∑|q|<Q(1−|q|Q)∑a<n,n+q≤bzn+q¯zn, |
where Q is any positive integer.
Proof. See Lemma 8.17 of [11].
Lemma 2.5. Let t be a non-integer, α∈(0,1) and H≥3. Then we have
e(−α{t})=∑|h|≤Hch(α)e(ht)+O(min(1,1H‖t‖)), |
where
ch(α)=1−e(−α)2πi(h+α). |
Proof. See Lemma 12 of [1].
Lemma 2.6. For any real number θ, we have
min(1,1H‖θ‖)=+∞∑h=−∞ahe(hθ), |
where
ah≪min(log2HH,1|h|,Hh2). |
Proof. See (3) of [10].
Lemma 2.7. Let f(x) be a real differentiable function such that f′(x) is monotonic and f′(x)≥m>0, or f′(x)≤−m<0, throughout the interval [a,b]. Then
∫bae(f(x))dx≪1m. |
Proof. See Lemma 4.2 of [23].
Lemma 2.8. Suppose that M>1,c>1,c∉Z and γ>0. Let A(M;c,γ) denote the number of solutions of the following inequalities
|nc1+nc2−nc3−nc4|<γ,M<n1,n2,n3,n4≤2M. |
Then we have
A(M;c,γ)≪(γM4−c+M2)Mε. |
Proof. See Theorem 2 of [20].
The central focus of this paper is the study of the sum
Γ=∑X2<p1,p2,p3,p4,p5≤X[pc1]+[pc2]+[pc3]+[pc4]+[pc5]=N(pi+2,P(z))=1i=1,2,3,4,5logp. |
In order to prove Theorem 1.1, we need only to show that Γ>0. By the trivial orthogonality relation given by
∫10e(αh)dα={1,if h=0,0,otherwise |
we can write Γ as
Γ=∑X2<p1,p2,p3,p4,p5≤X(pi+2,P(z))=1i=1,2,3,4,5(logp)∫1−τ−τe(([pc1]+[pc2]+[pc3]+[pc4]+[pc5]−N)α)dα. | (3.1) |
By the definition of Λi in Lemma 2.2, we can see that
Λi=∑d|(pi+2,P(z))μ(d)={1,if (pi+2,P(z))=1,0,otherwise. |
Then by Lemma 2.2 we find that
Γ=∑X2<p1,p2,p3,p4,p5≤X(logp)Λ1Λ2Λ3Λ4Λ5∫1−τ−τe(([pc1]+[pc2]+[pc3]+[pc4]+[pc5]−N)α)dα≥∑X2<p1,p2,p3,p4,p5≤X(logp)∫1−τ−τe(([pc1]+[pc2]+[pc3]+[pc4]+[pc5]−N)α)dα×(Λ−1Λ+2Λ+3Λ+4Λ+5+Λ+1Λ−2Λ+3Λ+4Λ+5+Λ+1Λ+2Λ−3Λ+4Λ+5+Λ+1Λ+2Λ+3Λ−4Λ+5=+Λ+1Λ+2Λ+3Λ+4Λ−5−4Λ+1Λ+2Λ+3Λ+4Λ+5)=Γ1+Γ2+Γ3+Γ4+Γ5−4Γ6, | (3.2) |
By the symmetric property, we have
Γ1=Γ2=Γ3=Γ4=Γ5=∑X2<p1,p2,p3,p4,p5≤X(logp)Λ−1Λ+2Λ+3Λ+4Λ+5×∫1−τ−τe(([pc1]+[pc2]+[pc3]+[pc4]+[pc5]−N)α)dα,Γ6=∑X2<p1,p2,p3,p4,p5≤X(logp)Λ+1Λ+2Λ+3Λ+4Λ+5×∫1−τ−τe(([pc1]+[pc2]+[pc3]+[pc4]+[pc5]−N)α)dα. |
Hence, by (3.1) and (3.2) we obtain
Γ≥5Γ1−4Γ6. | (3.3) |
Now define
L±(α)=∑X2<p≤X(logp)e([pc]α)∑d|(p+2,P(z))λ±(d)=∑d|P(z)λ±(d)∑μX<p≤Xd|p+2(logp)e([pc]α). | (3.4) |
Consider Γ1 first. By (3.4) we can derive that
Γ1=∫1−τ−τL−(α)L+(α)4e(−Nα)dα=Γ11+Γ12, | (3.5) |
where
Γ11=∫τ−τL−(α)L+(α)4e(−Nα)dα, | (3.6) |
Γ12=∫1−ττL−(α)L+(α)4e(−Nα)dα. | (3.7) |
Similarly, we have
Γ6=∫τ−τL+(α)5e(−Nα)dα+∫1−ττL+(α)5e(−Nα)dα=:Γ61+Γ62. | (3.8) |
Now combining (3.3), (3.5) and (3.8) we get
Γ≥5Γ11−4Γ61+(5Γ12−4Γ62). | (3.9) |
In the following sections, we shall prove that
5Γ11−4Γ61≫X5−clog5X,Γ12,Γ62≪X5−c−ε. |
In this section, we will give an asymptotic formula for the integrals Γ11 and Γ61 defined by (3.6) and (3.8), respectively. We consider the sum
L(α)=∑d≤Dλ(d)∑X2<p≤Xd|p+2(logp)e([pc]α), | (4.1) |
where λ(d) are real numbers satisfying
|λ(d)|≤1,λ(d)=0if2|dorμ(d)=0. | (4.2) |
Furthermore, we define
I(α)=∫XX2e(tcα)dt. | (4.3) |
Lemma 4.1. Let L(α) and I(α) be defined by (4.1) and (4.3), respectively. Suppose that ξ and δ satisfy the following conditions
ξ+7δ<2and3ξ+6δ<2. |
Then for |α|≤τ, we have
L(α)=∑d≤Dλ(d)φ(d)I(α)+O(XlogAX), |
where A>0 is a sufficiently large constant.
Proof. See Lemma 2.8 in [16].
Lemma 4.2. Let L(α) and I(α) be defined by (4.1) and (4.3), respectively. Then we have
(i)∫|α|≤τ|I(α)|4dα≪X4−clog4X,(ii)∫|α|≤τ|L(α)|4dα≪X4−clog10X,(iii)∫10|L(α)|4dα≪X2+3ε. |
Proof. By using the trivial estimate |I(α)|≪X and L(x)≪Xlog2X, (ⅰ) and (ⅱ) follow from Lemma 2.9 in [16]. For (ⅲ), we have
∫10|L(α)|4dα=∑di≤Di=1,2,3,4λ(d1)λ(d2)λ(d3)λ(d4)∑X2<p1,p2,p3,p4≤Xdi|pi+2,i=1,2,3,44∏i=1(logpi)×∫10e(([pc1]+[pc2]−[pc3]−[pc4])α)dα=∑X2<p1,p2,p3,p4≤X[pc1]+[pc2]=[pc3]+[pc4]4∏i=1(logpi)∑di≤D,di|pi+2i=1,2,3,4λ(d1)λ(d2)λ(d3)λ(d4)≪(log4X)∑X2<n1,n2,n3,n4≤X[nc1]+[nc2]=[nc3]+[nc4]τ(n1+2)τ(n2+2)τ(n3+2)τ(n4+2)≪Xε(log4X)∑X2<n1,n2,n3,n4≤X[nc1]+[nc2]=[nc3]+[nc4]1≪X2ε∑X2<n1,n2,n3,n4≤X|nc1+nc2−nc3−nc4|<41≪X3ε(4X4−c+X2)≪X2+3ε, |
where Lemma 2.8 is applied in the last step.
Let
M±(α)=∑d≤Dλ±(d)φ(d)I(α)=N±I(α). |
Then we can easily get the elementary estimate
M±(α)≪|I(α)|logX. | (4.4) |
By using Lemma 4.2 and (4.4) we find that
L−(α)L+(α)4−M−(α)M+(α)4=(L−(α)−M−(α))L+(α)4+(L+(α)−M+(α))M−(α)L+(α)3+(L+(α)−M+(α))M−(α)M+(α)L+(α)2+(L+(α)−M+(α))M−(α)M+(α)2L+(α)+(L+(α)−M+(α))M−(α)M+(α)3≪XlogAX(|L+(α)|4+|L+(α)|3|I(α)|logX+|L+(α)|2|I(α)|2log2X=+|L+(α)||I(α)|3log3X+|I(α)|4log4X). | (4.5) |
Let
Jτ=∫τ−τM−(α)M+(α)4e(−Nα)dα. | (4.6) |
Then we can derive from Lemma 4.2, (3.6), (4.5) and (4.6) that
Γ11−Jτ≪∫|α|≤τ|L−(α)L+(α)4−M−(α)M+(α)4|dα≪XlogA−4X(∫|α|≤τ|L+(α)|4dα+∫|α|≤τ|I(α)|4dα)≪X5−clogA−14X. | (4.7) |
Define
J=∫+∞−∞I(α)5e(−Nα)dα. | (4.8) |
Then by the argument used in Lemma 2.13 of [16], we have
J≫X5−c. | (4.9) |
With the help of Lemma 2.7 we can get that I(α)≪|α|−1X1−c. Hence, from (4.6) and (4.8) we find that
|N−(N+)4J−Jτ|≪(log5X)∫|α|>τ|I(α)|5dα≪(log5X)∫|α|>τ|α|−5X5−5cdα≪X5−c−4ξlog5X≪X5−c−ε. | (4.10) |
Now combining (4.7) and (4.10) we obtain
Γ11=N−(N+)4J+O(X5−clogA−14X). | (4.11) |
Similarly, we can prove that
Γ61=(N+)5J+O(X5−clogA−14X). | (4.12) |
In this section we shall consider the upper bound for the integrals Γ12 and Γ62 defined by (3.7) and (3.8), respectively. Define
T(α,X)=∑d≤D∑X2<n≤Xd|n+2e([nc]α). |
Lemma 5.1. For α∈(0,1), we have
T(α,X)≪X2c+1320+εD720+logXαXc−1. |
Proof. The proof is exactly the same as that of Lemma 2.7 in [16], where the exponential pair (κ,λ)=(19,1318) is used. One can see [16] for details.
Lemma 5.2. Let f(n) be a complex valued function defined on n∈(X2,X]. Then we have
∑X2<n≤XΛ(n)f(n)=S1−S2−S3, |
where
S1=∑k≤X13μ(k)∑X2k<ℓ≤Xk(logℓ)f(kℓ),S2=∑k≤X23c(k)∑X2k<ℓ≤Xkf(kℓ),S3=∑X13<k≤X23a(k)∑X2k<ℓ≤XkΛ(ℓ)f(kℓ), |
and where a(k),c(k) are real numbers satisfying
|a(k)|≤τ(k),|c(k)|≤logk. |
Proof. The proof can be found on page 112 of [26].
Lemma 5.3. Suppose that 2<c<990479. Let λ(d) be real numbers that satisfy (4.2) and L(α) be defined by (4.1). Then we have
supα∈(τ,1−τ)|L(α)|≪X32−c4−ε. |
Proof. From (4.1), it is easy to see that
L(α)=L1(α)+O(X12+ε), | (5.1) |
where
L1(α)=∑d≤Dλ(d)∑X2<n≤Xd|n+2Λ(n)e([nc]α). |
Hence by (5.1) we need only to show that the estimation
supα∈(τ,1−τ)|L1(α)|≪X32−c4−ε | (5.2) |
holds for 2<c<990479. Let H=X8479. Then, by using Lemma 2.5, we get
L1(α)=∑|h|≤Hch(α)∑X2<n≤XΛ(n)∑d≤Dd|n+2λ(d)e((h+α)nc)+O((logX)∑X2<n≤Xmin(1,1H‖nc‖)). | (5.3) |
By Lemmas 2.3 and 2.6 with the exponential pair (κ,λ)=(16,23), we obtain
(logX)∑X2<n≤Xmin(1,1H‖nc‖)=(logX)∑X2<n≤X+∞∑k=−∞ake(knc)≤(logX)+∞∑k=−∞|ak||∑X2<n≤Xe(knc)|≪(logX)(Xlog2HH+∑1≤k≤H1k((kXc)16X12+XkXc)=+∑k>HHk2((kXc)16X12+XkXc))≪(logX)(XH−1+H16Xc6+12+X1−c)≪X32−c4−ε. | (5.4) |
Next we consider the first term on the right-hand side of (5.3). We write it in the following form
S(α)=∑|h|≤Hch(α)∑X2<n≤XΛ(n)f(n), | (5.5) |
where
f(n)=∑d≤Dd|n+2λ(d)e((h+α)nc). |
By applying Lemma 5.2 we find that
S(α)=S1−S2−S3, | (5.6) |
where
S1=∑|h|≤Hch(α)∑k≤X13μ(k)∑X2k<ℓ≤Xk(logℓ)f(kℓ), | (5.7) |
S2=∑|h|≤Hch(α)∑k≤X23c(k)∑X2k<ℓ≤Xkf(kℓ), | (5.8) |
S3=∑|h|≤Hch(α)∑X13<k≤X23a(k)∑X2k<ℓ≤XkΛ(ℓ)f(kℓ), | (5.9) |
and |a(k)|≤τ(k),|c(k)|≤logk. Clearly, by (5.8) we can write S2 as
S2=S21+S22, | (5.10) |
where
S21=∑|h|≤Hch(α)∑k≤X13c(k)∑X2k<ℓ≤Xkf(kℓ),S22=∑|h|≤Hch(α)∑X13<k≤X23c(k)∑X2k<ℓ≤Xkf(kℓ). | (5.11) |
Therefore, by (5.6) and (5.10) we have
S(α)≪|S1|+|S21|+|S22|+|S3|. | (5.12) |
We consider the sum S21 defined by (5.11) first. We change the order of the summation to write it in the following form
S21=∑d≤Dλ(d)∑|h|≤Hch(α)∑k≤X13c(k)∑X2k<ℓ≤Xkd|kℓ+2e((h+α)(kℓ)c). |
Since λ(d)=0 for 2|d, from the condition d|kℓ+2 we have that (k,d)=1. Hence there exists an integer ℓ0 such that kℓ+2≡0(modd) is equivalent to ℓ≡ℓ0(modd), which means that ℓ=ℓ0+md for some integer m. Therefore, we get
S21=∑d≤Dλ(d)∑|h|≤Hch(α)∑k≤X13(k,d)=1c(k)∑X2kd−ℓ0d<m≤Xkd−ℓ0de((h+α)kc(ℓ0+md)c). | (5.13) |
By using Lemma 2.3 with the exponential pair (κ,λ)=A2BA2B(0,1)=(120,3340) we find that the sum over m in (5.13) is given by
≪(|h+α|kdXc−1)120(Xkd)3340+(|h+α|kdXc−1)−1≪|h+α|120Xc20+3140k−3140d−3140+|h+α|−1k−1d−1X1−c. | (5.14) |
Then from (4.2), (5.13) and (5.14) we can obtain
S21≪Xε(Xc20+1720D940H120+X1−c). | (5.15) |
For the sum S1 given by (5.7), we can apply partial summation to get rid of the log factor and then proceed as in the same process for S21 to get
S1≪Xε(Xc20+1720D940H120+X1−c). | (5.16) |
Now we consider the sum S3. By a splitting argument, we can decompose S3 into O(logX) sums of the following form
W(K)=∑|h|≤Hch(α)∑K<k<K1a(k)∑X2k<ℓ≤XkΛ(ℓ)∑d≤Dd|kℓ+2λ(d)e((h+α)(kℓ)c), | (5.17) |
where
K1≤2K,X13≤K<K1≤X23. | (5.18) |
We assume that K≥X12 first. It follows from (5.17), (5.18) and Cauchy's inequality that
|W(K)|2≪XεKmaxγ∈(τ,H+1)∑K<k<K1|∑X2k<ℓ≤XkΛ(ℓ)∑d≤Dd|kℓ+2λ(d)e(γ(kℓ)c)|2. | (5.19) |
Suppose that Q is an integer which satisfies
1≤Q≪XK. | (5.20) |
For the inner sum over ℓ in (5.19), by applying Lemma 2.4 we can derive that
|W(K)|2≪X1+εQmaxγ∈(τ,H+1)∑K<k<K1∑|q|≤Q(1−|q|Q)∑X2k<ℓ,ℓ+q≤XkΛ(ℓ)×∑d1≤Dd1|kℓ+2λ(d1)e(−γ(kℓ)c)Λ(ℓ+q)∑d2≤Dd2|k(ℓ+q)+2λ(d2)e(γ(k(ℓ+q))c)≪X1+εQmaxγ∈(τ,H+1)∑d1≤D∑d2≤Dλ(d1)λ(d2)∑|q|≤Q(1−|q|Q)×∑X2K1<ℓ,ℓ+q≤XKΛ(ℓ)Λ(ℓ+q)S, | (5.21) |
where
S=∑˜K<k≤~K1d1|kℓ+2d2|k(ℓ+q)+2e(γkc((ℓ+q)c−ℓc)) |
and
˜K=max(K,X2ℓ,X2(ℓ+q)),~K1=min(K1,Xℓ,Xℓ+q). |
Since λ(d)=0 for 2|d, we can assume that (d1d2,2)=1. Then it follows from d1|kℓ+2 and d2|k(ℓ+q)+2 that (d1,ℓ)=(d2,ℓ+q)=1. Hence there exists an integer k0 that is dependent on ℓ,h,d1,d2 such that the pair of conditions kℓ+2≡0(modd1) and k(ℓ+q)+2≡0(modd2) is equivalent to the congruence k≡k0(mod[d1,d2]). Thus, we have
S=∑˜K−k0[d1,d2]<m≤~K1−k0[d1,d2]e(F(m)), | (5.22) |
where
F(m)=γ(k0+m[d1,d2])c((ℓ+q)c−ℓc). | (5.23) |
For q=0, by the trivial estimate we have
S≪K[d1,d2]. | (5.24) |
For the case q≠0, by (5.23) we get
|F(j)(m)|≍|γ||q|ℓc−1[d1,d2]Kc−1(K[d1,d2])1−j,j≥1. |
We apply Lemma 2.3 with the following exponential pair
(κ,λ)=A(1384+ε,5584+ε)=(13194+ε,7697+ε) |
to derive that
S≪(|γ||q|ℓc−1[d1,d2]Kc−1)13194+ε(K[d1,d2])7697+ε+(|γ||q|ℓc−1[d1,d2]Kc−1)−1≪Xε(|γ|13194|q|13194ℓ13194(c−1)[d1,d2]−139194K13194c+139194+|γ|−1|q|−1ℓ1−c[d1,d2]−1K1−c). | (5.25) |
Note that
∑d1≤D∑d2≤D1[d1,d2]139194=∑r≤D∑d1≤D∑d2≤Dr=(d1,d2)(rd1d2)139194≪∑r≤D∑k1≤Dr∑k2≤Dr(1rk1k2)139194≪∑r≤Dr−139194(Dr)5597≪D5597, |
and
∑d1≤D∑d2≤D1[d1,d2]≪(logD)3. |
Then we use the above two estimates, (5.21), (5.24) and (5.25) to get
|W(K)|2≪X1+εQmaxγ∈(τ,H+1)∑X2K1<ℓ≤XKΛ(ℓ)2∑d1≤D∑d2≤DK[d1,d2]+X1+2εQmaxγ∈(τ,H+1)∑d1≤D∑d2≤D∑0<|q|<Q∑X2K1<ℓ,ℓ+q≤XKΛ(ℓ)Λ(ℓ+q)×(|γ|13194|q|13194ℓ13194(c−1)[d1,d2]−139194K13194c+139194+|γ|−1|q|−1ℓ1−c[d1,d2]−1K1−c)≪X2+εQ−1+X1+2εQ|γ0|13194K13194c+139194(∑d1≤D∑d2≤D1[d1,d2]139194)(∑0<|q|<Q|q|13194)×(∑X2K1<ℓ≤XKℓ13194(c−1))+X1+2εQ|γ0|−1K1−c(∑d1≤D∑d2≤D1[d1,d2])×(∑0<|q|<Q|q|−1)(∑X2K1<ℓ≤XKℓ1−c)≪Xε(X2Q−1+X13c+375194Q13194D5597|γ0|13194K−2197+X|γ0|−1Q−1K1−c) | (5.26) |
for some γ0∈[τ,H+1]. We choose
Q0=X13207(1−c)D−110207|γ0|−13207K1469,Q=[min(Q0,XK−1)]. |
Then, it is easy to check that
Q−1≍Q−10+KX−1. | (5.27) |
Substituting (5.27) into (5.26), we obtain
|W(K)|2≪Xε(X2(Q−10+KX−1)+X13c+375194Q13194D5597|γ0|13194K−2197=+X|γ0|−1(Q−10+KX−1)K1−c)≪Xε(X13c+380207D110207|γ0|13207+X53+X553−181c414D110207|γ0|−194207+X1−c2|γ0|−1), |
which implies that
|W(K)|≪Xε(X13c+380414D55207|γ0|13414+X56+X553−181c828D55207|γ0|−97207+X12−c4|γ0|−12). | (5.28) |
When K<X12, we can represent W(K) as follows:
W(K)=∑|h|≤Hch(α)∑X2K1<ℓ≤XKΛ(ℓ)∑max(K,X2ℓ)<k≤min(K1,Xℓ)a(k)×∑d≤Dd|kℓ+2λ(d)e((h+α)(kℓ)c). |
Now we have that XK≫X12, then, we may proceed as in (5.19)–(5.28) but with roles of k and ℓ reversed. Thus we can again derive the estimate (5.28). Consequently, we obtain
S3≪Xε(X13c+380414D55207|γ0|13414+X56+X553−181c828D55207|γ0|−97207+X12−c4|γ0|−12). | (5.29) |
To bound S22, we use the same methodology as for S3 to derive that
S22≪Xε(X13c+380414D55207|γ0|13414+X56+X553−181c828D55207|γ0|−97207+X12−c4|γ0|−12). | (5.30) |
Now combining (5.12), (5.15), (5.16), (5.29) and (5.30) and from the fact that γ0∈[τ,H+1], we find that
S(α)≪Xε(Xc20+1720D940H120+X1−c+X13c+380414D55207|γ0|13414+X56=+X553−181c828D55207|γ0|−97207+X12−c4|γ0|−12)≪Xε(Xc20+81519580+9δ40+X1−c+X13c414+1011811017+55δ207+X56=+X553828+c4+55δ207−97ξ207+X12+c4−ξ2). |
Therefore, from condition (2.1) we conclude that if 2<c<990479 then
supα∈(τ,1−τ)|S(α)|≪X32−c4−ε. | (5.31) |
With the help of (5.3)–(5.5) and (5.31), we finally obtain that
supα∈(τ,1−τ)|L1(α)|≪X32−c4−ε |
holds for 2<c<990479, and the proof of Lemma 5.3 is completed.
Lemma 5.4. Suppose that 2<c<990479. Then we have
∫1−ττ|L(α)|5dα≪X5−c−ε. |
Proof. Let G(α)=¯L(α)|L(α)|3. We have
|∫1−ττ|L(α)|5dα|=|∑d≤Dλ(d)∑X2<p≤Xd|p+2(logp)∫1−ττe([pc]α)G(α)dα|≤(logX)∑d≤D∑X2<p≤Xd|p+2|∫1−ττe([pc]α)G(α)dα|≤(logX)∑d≤D∑X2<n≤Xd|n+2|∫1−ττe([nc]α)G(α)dα|. | (5.32) |
From (5.32) and Cauchy's inequality, we get
|∫1−ττ|L(α)|5dα|2≪X(logX)3∑d≤D∑X2<n≤Xd|n+2|∫1−ττe([nc]α)G(α)dα|2=X(logX)3∫1−ττ¯G(β)dβ∫1−ττT(α−β,X)G(α)dα≪X(logX)3∫1−ττ|G(β)|dβ∫1−ττ|T(α−β,X)G(α)|dα. | (5.33) |
Now
∫1−ττ|T(α−β,X)G(α)|dα≪∫τ<α<1−τ|α−β|≤X−c|T(α−β,X)G(α)|dα+∫τ<α<1−τ|α−β|>X−c|T(α−β,X)G(α)|dα. | (5.34) |
By the trivial bound T(α,X)≪XlogX and Lemma 5.3, we have
∫τ<α<1−τ|α−β|≤X−c|T(α−β,X)G(α)|dα≪X(logX)supα∈(τ,1−τ)|G(α)|∫|α−β|≤X−cdα≪X1−c(logX)supα∈(τ,1−τ)|L(α)|4≪X7−2c−ε. | (5.35) |
From Lemmas 5.1 and 5.3, we obtain
∫τ<α<1−τ|α−β|>X−c|T(α−β,X)G(α)|dα≪∫τ<α<1−τ|α−β|>X−c|L(α)|4(X2c+1320+εD720+X1−clogX|α−β|)dα≪X2c+1320+εD720∫10|L(α)|4dα+X1−c(logX)supα∈(τ,1−τ)|L(α)|4∫|α−β|>X−c1|α−β|dα≪X2c+5320+7δ20+ε+X7−2c−ε≪X7−2c−ε, | (5.36) |
where (ⅲ) of Lemma 4.2 is used. It follows from (5.34)–(5.36) that
∫1−ττ|T(α−β,X)G(α)|dα≪X7−2c−ε. | (5.37) |
Combining (5.33), (5.37) and (ⅲ) of Lemma 4.3, we get
|∫1−ττ|L(α)|5dα|2≪X(logX)3X7−2c−ε∫10|L(α)|4dα≪X10−2c−ε2. | (5.38) |
Now Lemma 5.4 follows from (5.38).
We are now in a position to estimate Γ12 and Γ62. By Hölder's inequality and Lemma 5.4 we find that
|Γ12|≪∫1−ττ|L−(α)||L+(α)|4dα≪(∫1−ττ|L−(α)|5dα)15(∫1−ττ|L+(α)|5dα)45≪X5−c−ε. | (5.39) |
Similarly, for Γ62 we have
|Γ62|≪∫1−ττ|L+(α)|5dα≪X5−c−ε. | (5.40) |
Proposition 6.1. We have
5Γ11−4Γ61≫X5−clog5X. |
Proof. It follows from (4.11), (4.12) and Lemma 2.1(3) that
5Γ11−4Γ61=(5N−−4N+)(N+)4J+O(X5−clogA−14X)≥(5f(logDlogz)−4F(logDlogz))(1+O(log−1/3D))B5J+O(X5−clogA−14X)=(5f(134)−4F(134))B5J+O(X5−c−ε)=40eγ13(log94−45−45∫942log(t−1)tdt)B5J+O(X5−c−ε)≥0.001B5J+O(X5−c−ε)≫X5−clog5X, |
where the following trivial estimate is used:
B≍1logX. |
Now according to (3.9), (5.39), (5.40) and Proposition 6.1, we obtain
Γ≥(5Γ11−4Γ61)+O(|Γ12|+|Γ62|)≫X5−clog5X, |
which implies that Γ>0 for a sufficiently large natural number N. Then, (1.3) would have a solution in primes p1,p2,p3,p4,p5 satisfying
(p1+2,P(z))=(p2+2,P(z))=(p3+2,P(z))=(p4+2,P(z))=(p5+2,P(z))=1. | (6.1) |
Suppose that pi+2 has l prime factors, counted with multiplicity. From (6.1) and the condition X2<pi≤X we see that
X+2≥pi+2≥zl=Xηl. |
Then, l≤η−1. This means that pj+2 has at most [62273960−1916c] prime factors counted with multiplicity. Now Theorem 1.1 is proved.
The author declares that he has not used Artificial Intelligence (AI) tools in the creation of this article.
The author would like to thank the anonymous referees for many useful comments on the manuscript.
The author declares no conflict of interest.
[1] | K. Buriev, Additive problems with prime numbers, Ph.D. thesis, Moscow University, 1989. |
[2] |
R. Baker, Some Diophantine equations and inequalities with primes, Funct. Approx. Comment. Math., 64 (2021), 203–250. https://doi.org/10.7169/facm/1912 doi: 10.7169/facm/1912
![]() |
[3] |
J. Brüdern, E. Fouvry, Lagrange's Four Squares Theorem with almost prime variables, J. Reine Angew. Math., 454 (1994), 59–96. https://doi.org/10.1515/crll.1994.454.59 doi: 10.1515/crll.1994.454.59
![]() |
[4] | J. R. Chen, On the representation of a larger even integer as the sum of a prime and the product of at most two primes, Sci. Sinica, 16 (1973), 157–176. |
[5] |
Y. C. Cai, M. G. Lu, Additive problems involving primes of special type, Acta Arith., 140 (2009), 189–204. https://doi.org/10.4064/aa140-2-6 doi: 10.4064/aa140-2-6
![]() |
[6] |
Y. C. Cai, On a Diophantine equation involving primes, Ramanujan J., 50 (2019), 151–162. https://doi.org/10.1007/s11139-018-0027-6 doi: 10.1007/s11139-018-0027-6
![]() |
[7] | S. W. Graham, G. Kolesnik, Van der Corput's Method of Exponential Sums, New York: Cambridge University Press, 1991. https://doi.org/10.1017/cbo9780511661976 |
[8] | G. Greaves, Sieves in Number Theory, Berlin: Springer, 2001. https://doi.org/10.1007/978-3-662-04658-6 |
[9] |
L. K. Hua, Some results in additive prime number theory, Quart. J. Math. Oxford, 9 (1938), 68–80. https://doi.org/10.1093/qmath/os-9.1.68 doi: 10.1093/qmath/os-9.1.68
![]() |
[10] |
D. R. Heath-Brown, The Pjatecki∨i-Šapiro prime number theorem, J. Number Theory, 16 (1983), 242–266. https://doi.org/10.1016/0022-314X(83)90044-6 doi: 10.1016/0022-314X(83)90044-6
![]() |
[11] | H. Iwaniec, E. Kowalski, Analytic Number Theory, New York: American Mathematical Society, 2004. |
[12] |
A. Kumchev, T. Nedeva, On an equation with prime numbers, Acta Arith., 83 (1998), 117–126. https://doi.org/10.4064/aa-83-2-117-126 doi: 10.4064/aa-83-2-117-126
![]() |
[13] |
M. Laporta, D. I. Tolev, On an equation with prime numbers, Mat. Zametki, 57 (1995), 926–929. https://doi.org/10.1007/bf02304564 doi: 10.1007/bf02304564
![]() |
[14] | J. Li, M. Zhang, On a Diophantine inequality with five prime variables, preprint paper, 2018. |
[15] | J. Li, M. Zhang, On a Diophantine equation with three prime variables, Integers, 10 (2019), A39. |
[16] |
J. Li, F. Xue, M. Zhang, On a ternary Diophantine equality involving fractional powers with prime variables of a special form, Ramanujan J., 58 (2022), 1171–1199. https://doi.org/10.1007/s11139-021-00517-5 doi: 10.1007/s11139-021-00517-5
![]() |
[17] |
S. Li, On a Diophantine equation with prime numbers, Int. J. Number Theory, 15 (2019), 1601–1616. https://doi.org/10.1142/s1793042119300011 doi: 10.1142/s1793042119300011
![]() |
[18] |
K. Matomäki, X. Shao, Vinogradov's three primes theorem with almost twin primes, Compos. Math., 153 (2017), 1220–1256. https://doi.org/10.1112/s0010437x17007072 doi: 10.1112/s0010437x17007072
![]() |
[19] | Z. H. Petrov, On an equation involving fractional powers with prime numbers of a special type, God. Sofiĭ. Univ. "Sv. Kliment Okhridski." Fac. Mat. Inform., 104 (2017), 171–183. |
[20] |
O. Robert, P. Sargos, Three-dimensional exponential sums with monomials, J. Reine Angew. Math., 591 (2006), 1–20. https://doi.org/10.1515/crelle.2006.012 doi: 10.1515/crelle.2006.012
![]() |
[21] | B. I. Segal, A general theorem concerning some properties of an arithmetical function, C. R. Acad. Sci. URSS, 3 (1933), 95–98. |
[22] | B. I. Segal, The Waring theorem with fractional and irrational degrees, in Russian, Trudy Mat. Inst. Stekov, 5 (1933), 73–86. |
[23] | E. G. Titchmarsh, The Theory of the Riemann Zeta-Function, New York: Oxford University Press, 1986. |
[24] |
D. I. Tolev, Additive problems with prime numbers of special type, Acta Arith., 96 (2000), 53–88. https://doi.org/10.4064/aa96-1-2 doi: 10.4064/aa96-1-2
![]() |
[25] | I. M. Vinogradov, Representation of an odd number as a sum of three primes, C. R. (Doklady) Acad. Sci. URSS, 15 (1937), 291–294. |
[26] |
R. C. Vaughan, An elementary method in prime number theory, Acta Arith., 37 (1980), 111–115. https://doi.org/10.4064/aa-37-1-111-115 doi: 10.4064/aa-37-1-111-115
![]() |
[27] | W. G. Zhai, X. D. Cao, A Diophantine equation with prime numbers, in Chinese, Acta Math. Sin., 45 (2002), 443–454. |