In this paper, by using special integrals and integer powers of the arcsin function, we obtained the recurrences or explicit expressions of some parametric Euler-type series involving multiple harmonic sums and multiple t-harmonic sums. According to our results, these Euler-type series are all expressible in terms of π, ln(2), and zeta values. In particular, by specifying the parameters, we presented as examples the evaluations of some special series, including some known ones in the literature and some new ones.
Citation: Jiaye Lin. Evaluations of some Euler-type series via powers of the arcsin function[J]. AIMS Mathematics, 2025, 10(4): 8116-8130. doi: 10.3934/math.2025372
[1] | H. C. Vidya, B. A. Rao . Utilization of Ramanujan-type Eisenstein series in the generation of differential equations. AIMS Mathematics, 2024, 9(10): 28955-28969. doi: 10.3934/math.20241405 |
[2] | Ling Zhu . New inequalities of Wilker's type for hyperbolic functions. AIMS Mathematics, 2020, 5(1): 376-384. doi: 10.3934/math.2020025 |
[3] | Li Chen, Dmitry V. Dolgy, Taekyun Kim, Dae San Kim . Probabilistic type 2 Bernoulli and Euler polynomials. AIMS Mathematics, 2024, 9(6): 14312-14324. doi: 10.3934/math.2024696 |
[4] | Bai-Ni Guo, Dongkyu Lim, Feng Qi . Series expansions of powers of arcsine, closed forms for special values of Bell polynomials, and series representations of generalized logsine functions. AIMS Mathematics, 2021, 6(7): 7494-7517. doi: 10.3934/math.2021438 |
[5] | Kwang-Wu Chen, Fu-Yao Yang . Infinite series involving harmonic numbers and reciprocal of binomial coefficients. AIMS Mathematics, 2024, 9(7): 16885-16900. doi: 10.3934/math.2024820 |
[6] | Taekyun Kim, Hye Kyung Kim, Dae San Kim . Some identities on degenerate hyperbolic functions arising from $ p $-adic integrals on $ \mathbb{Z}_p $. AIMS Mathematics, 2023, 8(11): 25443-25453. doi: 10.3934/math.20231298 |
[7] | Chunli Li, Wenchang Chu . Remarkable series concerning $ \binom{3n}{n} $ and harmonic numbers in numerators. AIMS Mathematics, 2024, 9(7): 17234-17258. doi: 10.3934/math.2024837 |
[8] | Ling Zhu, Branko Malešević . Optimal bounds for the sine and hyperbolic tangent means by arithmetic and centroidal means in exponential type. AIMS Mathematics, 2021, 6(12): 13024-13040. doi: 10.3934/math.2021753 |
[9] | Long Li . Double series expansions for $ \pi $. AIMS Mathematics, 2021, 6(5): 5000-5007. doi: 10.3934/math.2021294 |
[10] | Zeinab S. I. Mansour, Maryam Al-Towailb . A $ q $-Type $ k $-Lidstone series for entire functions. AIMS Mathematics, 2023, 8(6): 13525-13542. doi: 10.3934/math.2023686 |
In this paper, by using special integrals and integer powers of the arcsin function, we obtained the recurrences or explicit expressions of some parametric Euler-type series involving multiple harmonic sums and multiple t-harmonic sums. According to our results, these Euler-type series are all expressible in terms of π, ln(2), and zeta values. In particular, by specifying the parameters, we presented as examples the evaluations of some special series, including some known ones in the literature and some new ones.
Generalized harmonic numbers and odd harmonic numbers are defined by
H(s)n:=n∑k=11ksandO(s)n:=n∑k=11(2k−1)s, |
for n∈N0 and s∈N. Multiple harmonic sums (MHSs) ζn(s) and multiple t-harmonic sums (MtSs) tn(s) are defined by
ζn(s):=∑n≥n1>⋯>nk≥11ns11⋯nskkandtn(s):=∑n≥n1>⋯>nk≥1k∏j=11(2nj−1)sj, | (1.1) |
for n∈N0 and s=(s1,…,sk)∈Nk. By convention, ζn(∅)=tn(∅):=1 for n≥0, and ζn(s)=tn(s):=0 for n<k. When k=1, the MHSs and MtSs reduce respectively, to harmonic numbers and odd harmonic numbers, i.e., ζn(s)=H(s)n and tn(s)=O(s)n. Moreover, for s1>1, define the multiple zeta values (MZVs) and multiple t-values (MtVs) by
ζ(s):=∑n1>⋯>nk≥11ns11⋯nskkandt(s):=∑n1>⋯>nk≥1k∏j=11(2nj−1)sj, | (1.2) |
which can be viewed as the limits of MHSs and MtSs, respectively. The concept of MZVs was introduced by Hoffman [1] and Zagier [2] in the early 1990s, and the concept of MtVs was formally introduced by Hoffman [3] in 2019. Similarly to MZVs and MtVs, let s=(s1,…,sk)∈Nk and σ=(σ1,…,σk)∈{±1}k with (s1,σ1)≠(1,1), and we define the alternating multiple zeta values (AMZVs) by
ζ(s;σ):=∑n1>⋯>nk≥1σn11⋯σnkkns11⋯nskk. |
Summation formulas on the (alternating) multiple zeta values have attracted extensive attention, and the readers are referred to the papers due to, e.g., Alegri et al. [4], Hoffman [1,5], and Shen and Cai [6].
The linear Euler sums Sp,q:=∑∞n=1H(p)n/nq were first considered by Euler in correspondence with Goldbach in 1742 (see [7, p. 253]). Euler discovered that the linear sums Sp,q can be evaluated in terms of zeta values when p=1, p=q, p+q is odd, and (p,q)∈{(2,4),(4,2)}. The nonlinear Euler sums were introduced by Flajolet and Salvy [8] in 1998. These sums are infinite series of the form
Sp1p2…pk,q:=∞∑n=1H(p1)nH(p2)n⋯H(pk)nnq, |
where p1,p2,…,pk,q∈N, with p1≤p2≤⋯≤pk and q≥2. Flajolet and Salvy used the contour integral representations and residue computation to establish explicit formulas of several classes of Euler sums. In 2020, by permutations and compositions, Xu and Wang[9] established explicit formulas of Euler sums Sp1p2…pk,q and various alternating Euler sums, and showed that these infinite series are expressible in terms of MZVs or AMZVs, respectively. More classical results on Euler sums can be found in, e.g., the works due to Bailey et al. [10], Borwein et al. [11], Chu [12,13], Wang and Lyu [14], Xu[15], Xu et al. [16], and Zheng [17].
Recently, various Euler-type series have been investigated, including Euler sums involving negative powers of two [18], Euler-type series on odd harmonic numbers [19], and Euler-type series on hyperharmonic numbers [20,21]. Additionally, in 2017, Hoffman [22] established some general series identities by symmetric functions and MZVs. For example, he showed that
∞∑n=1Pk(Hn,H(2)n,…,H(k)n)n2=k+32ζ(k+2)−12k∑j=2ζ(j)ζ(k+2−j),∞∑n=0Pk(Hn,H(2)n,…,H(k)n)(n+1)(n+2)⋯(n+q)=1(q−1)!(q−1)k+1, |
for integers k≥0 and q≥2. According to the definition of the multivariate polynomials Pk, it can be verified that ζn({1}k)=Pk(Hn,H(2)n,…,H(k)n), where {⋯}k denotes k repetitions of a substring, with {⋯}0=∅. Therefore, Hoffman's results can also be viewed as Euler-type series identities. Kuba and Panholzer [23], and Ma and Wang [24], further generalized Hoffman's results, and established more infinite series identities on Euler-type series involving ζn({1}k) and ζ⋆n({1}k), where the star version ζ⋆n(s) was obtained from the definition of ζn(s) by replacing n>n1>⋯>nk≥1 by n≥n1≥⋯≥nk≥1. More recent works on various Euler-type series can be found in the papers of Campbell and Chen [25], Chen and Wang [26], Liu and Wang [27], and Liu et al. [28]. Note that infinite series involving (odd) harmonic numbers and central binomial coefficients have also attracted extensive attention recently. For example, Campbell et al.[29] used the theory of colored multiple zeta values and various special integrals to evaluate this kind of infinite series, and Nimbran et al.[30] used the series expansions of integer powers of the arcsin function to determine the values of such series. More recent results of this type can be found in, e.g., the papers of Wang and Xu[31], Wang and Yuan[32], and Wei and Xu[33].
In this paper, we will use the series expansions of arcsinn(x) as well as some integrals of trigonometric functions and inverse trigonometric functions to obtain the explicit evaluations of more Euler-type series. For n=1,2,3, the series expansions of arcsinn(x) were given in Edwards' textbook [34, pp. 77-82], and the expansion of arcsin4(x) was recorded in Ramanujan's notebooks [7, Proposition 15]. In the classical book [35, Eq (46)] of Schwatt, a general formula for the series expansion of arcsinn(x), for n∈N, was presented. However, two elegant formulas of arcsinn(x) were presented by Borwein and Chamberland [36, Eqs (1.1) and (1.3)], which can be reformulated as
arcsin2p+1(x)(2p+1)!=∞∑n=p(2nn)4ntn({2}p)(2n+1)x2n+1,forp≥0, | (1.3) |
arcsin2p(x)(2p)!=∞∑n=p4n−p(2nn)ζn−1({2}p−1)n2x2n,forp≥1. | (1.4) |
By means of the expansions (1.3) and (1.4), the following parametric series involving MHSs ζn({2}p) and MtSs tn({2}p) will be studied in this paper:
∞∑n=pζn({2}p)(n+1)(n+m+1),∞∑n=ptn({2}p)(2n+1)(2n+2m+1), | (1.5) |
∞∑n=pζn({2}p)(2n+k)(2n+k+1),∞∑n=ptn({2}p)(2n+k−1)(2n+k), | (1.6) |
for m∈N, p∈N0, and k=0,1,2. Note that according to our previous results [37, Lemma 2.1], ζn({r}k) and tn({r}k) can be expressed in terms of harmonic numbers and odd harmonic numbers by the complete Bell polynomials:
ζn({r}k)=1k!Yk(H(r)n,−1!H(2r)n,…,(−1)k−1(k−1)!H(kr)n),tn({r}k)=1k!Yk(O(r)n,−1!O(2r)n,…,(−1)k−1(k−1)!O(kr)n). |
For example,
ζn(r)=H(r)n,ζn(r,r)=12{(H(r)n)2−H(2r)n},tn(r)=O(r)n,tn(r,r)=12{(O(r)n)2−O(2r)n}. |
Therefore, the parametric series in (1.5) and (1.6) are indeed Euler-type. We establish in Section 2 the recurrences of the series in (1.5) by the integral ∫π20θnsin2m(θ)dθ, and derive in Sections 3–5 the explicit expressions of the series in (1.6) by the integrals ∫θ0arcsinn(x)/x3dx, ∫π20θn/sin2(θ)dθ, and ∫θ0tncos(t)dt. The results show that the above general Euler-type series are all expressible in terms of powers of π, ln(2), and zeta values. By specifying the parameters, we present as examples the evaluations of some special Euler-type series, including some known ones appearing in recent papers due to Hoffman [22], Campbell and Chen [25], and Chen and Wang [26].
In this section, we consider two kinds of infinite series Z(p)m and T(p)m, defined by
Z(p)m:=∞∑n=pζn({2}p)(n+1)(n+m+1)andT(p)m:=∞∑n=ptn({2}p)(2n+1)(2n+2m+1), |
for integers m≥1 and p≥0. These series are closely related to the integral
I(n,m):=1n!∫π20θnsin2m(θ)dθ,forn,m≥0. |
According to Orr's result [38, Eq (14)], the following expression for I(n,m) holds:
I(n,m)=πn+1(2mm)22m+n+1(n+1)!−m∑k=1⌊n−12⌋∑j=1(−1)j(2mm+k)πn−2j+122m+nk2j(n+1−2j)!+δ⌊n+12⌋,n+12m∑k=1(−1)k+⌊n+12⌋(2mm+k)22m+nkn+1. |
In the next two theorems, we will show that by means of the integral I(n,m), the recurrences of the series Z(p)m and T(p)m can be established.
Theorem 2.1. For integers m≥1 and p≥0, the series Z(p)m satisfy the following recurrence:
m∑k=1a(m,k)kZ(p)k=(2mm)π2p+222m−1(2p+3)!−4p+2πI(2p+2,m), | (2.1) |
where
a(m,k)=−2k4m(2m−2km−k)(2kk). |
Therefore, the series Z(p)m is expressible in terms of powers of π.
Proof. Applying a change of variable x→sin(θ) to Eq (1.4) gives
θ2p(2p)!=∞∑n=p4n−p(2nn)ζn−1({2}p−1)n2sin2n(θ). | (2.2) |
Multiply (2.2) with sin2m(θ), integrate from 0 to π2, and use Wallis' formula. Then we obtain
I(2p,m)=∞∑n=p4n−p(2nn)ζn−1({2}p−1)n2∫π20sin2n+2m(θ)dθ=π2∞∑n=p(2n+2mn+m)(2nn)ζn−1({2}p−1)4p+mn2=π2∞∑n=pζn−1({2}p−1)4pn2m∏k=1(1−12n+2k). |
Define f2n(m)=∏mk=1(1−12n+2k). Based on one of our previous results [37, Lemma 4.1], fl(m) can be expanded as fl(m)=1+∑mk=1a(m,k)l+2k, with a(m,k)=−2k4m(2m−2km−k)(2kk). Substituting this expansion back, and using the definition of MZVs as well as the method of partial fraction decomposition, we have
I(2p,m)=π2∞∑n=pζn−1({2}p−1)4pn2(1+m∑k=1a(m,k)2n+2k)=π2⋅4p∞∑n=pζn−1({2}p−1)n2+π4p+1∞∑n=pm∑k=1a(m,k)kζn−1({2}p−1)(1n2−1n(n+k))=πζ({2}p)2⋅4p+πζ({2}p)4p+1m∑k=1a(m,k)k−π4p+1m∑k=1a(m,k)kZ(p−1)k. |
The summation in the second term can be computed by Gould [39, Eq (3.90)]:
m∑k=1a(m,k)k=−24mm∑k=1(2m−2km−k)(2kk)=2(2mm)4m−2, | (2.3) |
which, together with the fact that ζ({2}p)=π2p(2p+1)! (see [40, Proposition 2.3]), gives the desired recurrence.
Example 2.1. When m=1, Eq (2.1) gives the following explicit expression:
∞∑n=pζn({2}p)(n+1)(n+2)=p+1∑j=1(−1)j+1π2p+2−2j(2p+3−2j)!. | (2.4) |
Setting p=1,2 in the above expression yields
∞∑n=0H(2)n(n+1)(n+2)=π26−1,∞∑n=0ζn(2,2)(n+1)(n+2)=π4120−π26+1. | (2.5) |
Note that Eq (2.5) was presented by Hoffman [22, Corollary 1]. Similarly, when m=2,3, setting p=1,2 gives
∞∑n=0H(2)n(n+1)(n+3)=π28−1116, | (2.6) |
∞∑n=0ζn(2,2)(n+1)(n+3)=π4160−11π296+4364,∞∑n=0ζn(2,2)(n+1)(n+4)=11π42160−341π23888+1181323328, | (2.7) |
where Eqs (2.6) and (2.7) can be found in Sofo [41, Lemma 1.1].
In the next theorem, we derive the recurrence of the series T(p)m.
Theorem 2.2. For integers m≥1 and p≥0, the series T(p)m satisfies the following recurrence:
m∑k=1a(m,k)kT(p)k=(2mm)(π2)2p+222m−1(2p+2)!−2I(2p+1,m), | (2.8) |
where a(m,k)=−2k4m(2m−2km−k)(2kk). Therefore, the series T(p)m is expressible in terms of powers of π.
Proof. Applying a change of variable x→sin(θ) to Eq (1.3), the series expansion turns into
θ2p+1(2p+1)!=∞∑n=p(2nn)4ntn({2}p)(2n+1)sin2n+1(θ). | (2.9) |
Multiply (2.9) with sin2m(θ), integrate from 0 to π2, and use Wallis' formula as well as the expansion of fl(m). Then we obtain
I(2p+1,m)=∞∑n=p(2nn)tn({2}p)4n(2n+1)∫π20sin2n+2m+1(θ)dθ=4m∞∑n=p(2nn)tn({2}p)(2n+2mn+m)(2n+1)(2n+2m+1)=∞∑n=ptn({2}p)(2n+1)2f2n+1(m)=∞∑n=ptn({2}p)(2n+1)2+m∑k=1a(m,k)∞∑n=ptn({2}p)(2n+1)2(2n+2k+1). | (2.10) |
By partial fraction decomposition and the definition of MtVs, we have
∞∑n=ptn({2})p(2n+1)2=t({2}p+1),∞∑n=ptn({2}p)(2n+1)2(2n+2k+1)=12k(t({2}p+1)−T(p)k). |
Setting m=0 in (2.10) gives t({2}p)=1(2p)!(π2)2p (see also [42, Lemma 1]). Combining the above results with (2.3) yields Eq (2.8).
Example 2.2. When m=1, Eq (2.8) can be reformulated as
∞∑n=0tn({2}p)(2n+1)(2n+3)=p∑j=1(−1)jπ2p+2−2j4p+1(2p+2−2j)!+(−1)p4p+1. |
In particular, setting p=1,2 yields
∞∑n=0O(2)n(2n+1)(2n+3)=π232−18,∞∑n=0tn(2,2)(2n+1)(2n+3)=π41536−π2128+132. | (2.11) |
Equation (2.11) was presented by Chen and Wang [26, Example 4.6]. Similarly, setting m=2, p=1,2 and m=3, p=1,2 in Eq (2.8) gives
∞∑n=0O(2)n(2n+1)(2n+5)=3π2128−112,∞∑n=0tn(2,2)(2n+1)(2n+5)=π42048−11π22048+148,∞∑n=0O(2)n(2n+1)(2n+7)=11π2576−2033240,∞∑n=0tn(2,2)(2n+1)(2n+7)=11π427648−341π282944+1823116640. |
Based on Eqs (1.3) and (1.4) as well as the integral ∫sin(θ)0arcsinn(x)x3dx, we can establish the expressions of another two kinds of infinite series.
Theorem 3.1. For an integer p≥0, the following series are expressible in terms of powers of π, ln(2), and zeta values:
∞∑n=pζn({2}p)2n(2n+1)=p∑k=1(−1)k(2k+1)(4−k−1)π2pζ(2k+1)(2p−2k)!+(−1)pζ(2p+1)−πln(2)(2p+1)! | (3.1) |
and
∞∑n=ptn({2}p)2n(2n−1)=p∑k=1(−1)k(2k+1)(1−4k)(π2)2p−2kζ(2k+1)(2p−2k)!16k+(−1)p−1(2p+1)ζ(2p+1)4p−(π2)2pln(2)(2p)!. | (3.2) |
Proof. Consider the integral 1(2p)!∫sin(θ)0arcsin2p(x)x3dx. On the one hand, by integration by parts twice, we have
1(2p)!∫sin(θ)0arcsin2p(x)x3dx=−θ2p2(2p)!sin2(θ)−θ2p−1cot(θ)2(2p−1)!+12(2p−2)!∫sin(θ)0arcsin2p−2(x)xdx. |
On the other hand, expanding arcsin2p(x) by (1.4), using partial fraction decomposition, and applying Eq (2.2), the integral on the left can be rewritten as
1(2p)!∫sin(θ)0arcsin2p(x)x3dx=∞∑n=p4n−p(2nn)ζn−1({2}p−1)sin2n−2(θ)n2(2n−2)=∞∑n=p4n−p(2nn)(ζn−1({2}p−1)sin2n−2(θ)n(2n−2)−ζn−1({2}p−1)sin2n−2(θ)2n2)=∞∑n=p4n−p(2nn)ζn−1({2}p−1)sin2n−2(θ)n(2n−2)−θ2p2(2p)!sin2(θ). |
Combining the above two expressions gives
∞∑n=p4n−p(2nn)ζn−1({2}p−1)sin2n−2(θ)n(2n−2)=12(2p−2)!∫sin(θ)0arcsin2p−2(x)xdx−θ2p−1cot(θ)2(2p−1)!. | (3.3) |
Now, integrate both sides of Eq (3.3) from 0 to π2. Then by Wallis' formula, the left-hand side of (3.3) turns into
∞∑n=p4n−p(2nn)ζn−1({2}p−1)n(2n−2)∫π20sin2n−2(θ)dθ=π4p∞∑n=pζn−1({2}p−1)(2n−2)(2n−1). |
The integral of the first term on the right-hand side of (3.3) can be computed by the following formula (see [37, Eq (5.2)]):
ζ(3,{2}p)=4p+2(2p+2)!π∫π20dθ∫sin(θ)0arcsin2p+2(x)xdx=p+1∑n=1(−1)nπ2p+2−2n(1−22n)(2p+3−2n)!22n−2nζ(2n+1)+(−1)p+12ζ(2p+3), |
and the integral of the second term can be evaluated by Orr's result [43, Eq (2.2)]:
∫π20θncot(θ)dθ=(π2)nln(2)+⌊n2⌋∑k=1(−1)kn!(1−4−k)πn−2kζ(2k+1)(n−2k)!2n+δ⌊n2⌋,n2(−1)n2n!ζ(n+1)2n. | (3.4) |
Substituting the above formulas into (3.3) gives (3.1).
Similarly, expanding the integral 1(2p+1)!∫sin(θ)0arcsin2p+1(x)x3dx in two different ways yields
∞∑n=p(2nn)4ntn({2}p)sin2n−1(θ)2n−1=1(2p−1)!∫sin(θ)0arcsin(x)2p−1xdx−θ2pcot(θ)(2p)! | (3.5) |
(see also [37, Theorem 3.7]). Integrate both sides of (3.5) from 0 to π2. Using Wallis' formula for the left, and the integral expression of t(3,{2}p) [37, Eq (5.1)] as well as the integral (3.4) for the right, we obtain Eq (3.2).
Example 3.1. Setting p=1,2,3 in Eq (3.1) gives
∞∑n=1H(2)n2n(2n+1)=−π2ln(2)6+5ζ(3)4,∞∑n=1ζn(2,2)2n(2n+1)=−π4ln(2)120+3π2ζ(3)8−59ζ(5)16,∞∑n=1ζn(2,2,2)2n(2n+1)=−π6ln(2)5040+3π4ζ(3)160−25π2ζ(5)32+377ζ(7)64. |
Setting p=1,2,3 in Eq (3.2) yields
∞∑n=1O(2)n2n(2n−1)=−π2ln(2)8+21ζ(3)16,∞∑n=1tn(2,2)2n(2n−1)=−π4ln(2)384+9π2ζ(3)128−155ζ(5)256,∞∑n=1tn(2,2,2)2n(2n−1)=−π6ln(2)46080+3π4ζ(3)2048−75π2ζ(5)2048+889ζ(7)4096. |
In this section, we establish the expressions of two Euler-type series by the integral of θn/sin2(θ).
Theorem 4.1. For an integer p≥0, the following series are expressible in terms of powers of π, ln(2), and zeta values:
∞∑n=pζn({2}p)(2n+1)(2n+2)=π2pln(2)(2p+1)!+p∑k=1(−1)k(1−4−k)π2p−2kζ(2k+1)(2p+1−2k)! | (4.1) |
and
∞∑n=ptn({2}p)2n(2n+1)=p∑k=1(−1)k(4k−1)(π2)2p−2kζ(2k+1)(2p−2k)!16k+(π2)2pln(2)(2p)!+(−1)pζ(2p+1)4p. | (4.2) |
Proof. Consider the integral ∫π20θnsin2(θ)dθ. On the one hand, by integration by parts, we have
∫π20θnsin2(θ)dθ=∫π20θnd(−cot(θ))=n∫π20θn−1cot(θ)dθ, | (4.3) |
which can be computed by Eq (3.4). On the other hand, expanding θ2p by Eq (2.2) and using Wallis' formula, we obtain
∫π20θ2psin2(θ)dθ=π2⋅(2p)!4p−1∞∑n=pζn−1({2}p−1)2n(2n−1), |
which, together with (4.3), gives Eq (4.1). Similarly, expanding θ2p+1 by Eq (2.9), using Wallis' formula, and applying Eq (4.3), we obtain Eq (4.2).
Example 4.1. Setting p=1,2,3 in Eq (4.1) gives
∞∑n=0H(2)n(2n+1)(2n+2)=π2ln(2)6−3ζ(3)4,∞∑n=0ζn(2,2)(2n+1)(2n+2)=π4ln(2)120−π2ζ(3)8+15ζ(5)16,∞∑n=0ζn(2,2,2)(2n+1)(2n+2)=π6ln(2)5040−π4ζ(3)160+5π2ζ(5)32−63ζ(7)64. | (4.4) |
Here, Eq (4.4) can be found in Campbell and Chen [25, p. 15], and Chen and Wang [26, Example 4.1]. Setting p=1,2,3 in Eq (4.2) yields
∞∑n=1O(2)n2n(2n+1)=π2ln(2)8−7ζ(3)16,∞∑n=1tn(2,2)2n(2n+1)=π4ln(2)384−3π2ζ(3)128+31ζ(5)256,∞∑n=1tn(2,2,2)2n(2n+1)=π6ln(2)46090−π4ζ(3)2048+15π2ζ(5)2048−127ζ(7)4097. |
In this section, we establish two series identities based on Eqs (2.2) and (2.9), which give the evaluations of the following two Euler-type series.
Theorem 5.1. For an integer p≥0, the following series are expressible in terms of powers of π, ln(2), and zeta values:
∞∑n=pζn({2}p)(2n+2)(2n+3)=p∑k=0p−k∑j=1(−1)k+j(4k−j−4k)π2p−2k−2jζ(2j+1)(2p+1−2k−2j)!+(1−ln(2))p∑k=0(−4)kπ2p−2k(2p+1−2k)! | (5.1) |
and
∞∑n=ptn({2}p)(2n+1)(2n+2)=(−1)pln(2)+(1−ln(2))p−1∑k=0(−1)k(π2)2p−2k(2p−2k)!−(−1)pp∑k=1ζ(2k+1)4k+p−1∑k=0p−k∑j=1(−1)k+j(1−4j)(π2)2p−2k−2jζ(2j+1)(2p−2k−2j)!16j. | (5.2) |
Proof. On the one hand, by integration by parts, we have
1(2p)!∫θ0t2pcos(t)dt=p∑k=0(−1)kθ2p−2ksin(θ)(2p−2k)!+p−1∑k=0(−1)kθ2p−2j−1cos(θ)(2p−2k−1)!. |
On the other hand, replace θ by t in (2.2), multiply both sides by cos(t), and integrate from 0 to θ. Then we have
1(2p)!∫θ0t2pcos(t)dt=∞∑n=p4n−p(2nn)ζn−1({2}p−1)n2(2n+1)sin2n+1(θ)=∞∑n=p4n−p(2nn)ζn−1({2}p−1)n2sin2n+1(θ)−2∞∑n=p4n−p(2nn)ζn−1({2}p−1)n(2n+1)sin2n+1(θ)=θ2psin(θ)(2p)!−2∞∑n=p4n−p(2nn)ζn−1({2}p−1)n(2n+1)sin2n+1(θ), |
where the last two steps can be obtained by partial fraction decomposition and Eq (2.9). Combining the above two results gives us an identity. Divide both sides of this identity by sin(θ), and integrate from 0 to π2. Then we have
2∞∑n=p4n−p(2nn)ζn−1({2}p−1)n(2n+1)∫π20sin2n(θ)dθ=p∑k=1(−1)k+1(2p−2k)!∫π20θ2p−2kdθ+p−1∑k=0(−1)k+1(2p−2k−1)!∫π20θ2p−2k−1cot(θ)dθ. |
Using Wallis' formula for the left, and applying the integral (3.4) for the second term on the right, we obtain Eq (5.1). Similarly, by means of the integral ∫π20t2p+1cos(t)dt, we have
∞∑n=p(2nn)4ntn({2}p)sin2n+2(θ)2n+2=p∑k=1(−1)k+1θ2p−2k+1sin(θ)(2p−2k+1)!+p∑k=0(−1)k+1θ2p−2kcos(θ)(2p−2k)!+(−1)p | (5.3) |
(see also [37, Theorem 3.8]). Divide (5.3) by sin(θ) and integrate from 0 to π2. The integral of the left-hand side is easy to calculate by Wallis' formula. The right-hand side turns into
p∑k=1(−1)k+1(2p−2k+1)!∫π20θ2p−2k+1dθ+p−1∑k=0(−1)k+1(2p−2k)!∫π20θ2p−2kcot(θ)dθ+(−1)p+1∫π20cos(θ)−1sin(θ)dθ. |
The first and third terms can be computed directly. For the second term, use the integral (3.4). Finally, we obtain Eq (5.2).
Example 5.1. Setting p=1,2 in Eq (5.1) gives
∞∑n=0H(2)n(2n+2)(2n+3)=−π2ln(2)6+π26+3ζ(3)4+4ln(2)−4,∞∑n=0ζn(2,2)(2n+2)(2n+3)=−π4ln(2)120+π4120−15ζ(5)16+π2ζ(3)8−2π23−3ζ(3)+2π2ln(2)3−16ln(2)+16, | (5.4) |
where Eq (5.4) can be found in Chen and Wang [26, Example 4.2]. Setting p=1,2 in Eq (5.2) yields
∞∑n=0O(2)n(2n+1)(2n+2)=−π2ln(2)8+π28+7ζ(3)16−ln(2),∞∑n=0tn(2,2)(2n+1)(2n+2)=−π4ln(2)384+π4384−7ζ(3)16+3π2ζ(3)128+π2ln(2)8−π28−31ζ(5)256+ln(2). |
In this paper, by using integer powers of the arcsin function and special integrals, we obtained the explicit expressions or recurrences of some parametric Euler-type series:
∞∑n=pζn({2}p)(n+1)(n+m+1),∞∑n=ptn({2}p)(2n+1)(2n+2m+1),∞∑n=pζn({2}p)(2n+k)(2n+k+1),∞∑n=ptn({2}p)(2n+k−1)(2n+k), |
for m∈N, p∈N0, and k=0,1,2. These Euler-type series are all expressible in terms of π, ln(2), and zeta values.
The author declares he (she) have not used Artificial Intelligence (AI) tools in the creation of this article.
The author would like to thank the anonymous referees for their valuable comments and suggestions and Professor Weiping Wang for his valuable discussions and kind help.
The author declares there is no conflict of interest.
[1] |
M. E. Hoffman, Multiple harmonic series, Pacific J. Math., 152 (1992), 275–290. https://doi.org/10.2140/PJM.1992.152.275 doi: 10.2140/PJM.1992.152.275
![]() |
[2] | D. Zagier, Values of zeta functions and their applications, In: A. Joseph, F. Mignot, F. Murat, B. Prum, R. Rentschler, First European Congress of Mathematics Paris, July 6–10, 1992, Progress in Mathematics, Birkhäuser Basel, 120 (1994), 497–512. https://doi.org/10.1007/978-3-0348-9112-7_23 |
[3] |
M. E. Hoffman, An odd variant of multiple zeta values, Commun. Number Theory Phys., 13 (2019), 529–567. https://doi.org/10.4310/CNTP.2019.v13.n3.a2 doi: 10.4310/CNTP.2019.v13.n3.a2
![]() |
[4] |
M. Alegri, J. L. López-Bonilla, P. S. K. Reddy, Results on Riemann zeta functions using compositions, J. Interdiscip. Math., 27 (2024), 1437–1447. https://doi.org/10.47974/JIM-1903 doi: 10.47974/JIM-1903
![]() |
[5] |
M. E. Hoffman, On multiple zeta values of even arguments, Int. J. Number Theory, 13 (2017), 705–716. https://doi.org/10.1142/S179304211750035X doi: 10.1142/S179304211750035X
![]() |
[6] |
Z. Shen, T. Cai, Some identities for multiple zeta values, J. Number Theory, 132 (2012), 314–323. https://doi.org/10.1016/j.jnt.2011.06.011 doi: 10.1016/j.jnt.2011.06.011
![]() |
[7] | B. C. Berndt, Ramanujan's notebooks, Part Ⅰ, New York: Springer-Verlag, 1985. https://doi.org/10.1007/978-1-4612-1088-7 |
[8] |
P. Flajolet, B. Salvy, Euler sums and contour integral representations, Exp. Math., 7 (1998), 15–35. https://doi.org/10.1080/10586458.1998.10504356 doi: 10.1080/10586458.1998.10504356
![]() |
[9] |
C. Xu, W. Wang, Explicit formulas of Euler sums via multiple zeta values, J. Symb. Comput., 101 (2020), 109–127. https://doi.org/10.1016/j.jsc.2019.06.009 doi: 10.1016/j.jsc.2019.06.009
![]() |
[10] |
D. H. Bailey, J. M. Borwein, R. Girgensohn, Experimental evaluation of Euler sums, Exp. Math., 3 (1994), 17–30. https://doi.org/10.1080/10586458.1994.10504573 doi: 10.1080/10586458.1994.10504573
![]() |
[11] |
D. Borwein, J. M. Borwein, R. Girgensohn, Explicit evaluation of Euler sums, Proc. Edinburgh Math. Soc., 38 (1995), 277–294. https://doi.org/10.1017/S0013091500019088 doi: 10.1017/S0013091500019088
![]() |
[12] |
W. Chu, Hypergeometric series and the Riemann zeta function, Acta Arith., 82 (1997), 103–118. https://doi.org/10.4064/AA-82-2-103-118 doi: 10.4064/AA-82-2-103-118
![]() |
[13] |
W. Chu, Infinite series identities on harmonic numbers, Results Math., 61 (2012), 209–221. https://doi.org/10.1007/s00025-010-0089-2 doi: 10.1007/s00025-010-0089-2
![]() |
[14] |
W. Wang, Y. Lyu, Euler sums and Stirling sums, J. Number Theory, 185 (2018), 160–193. https://doi.org/10.1016/j.jnt.2017.08.037 doi: 10.1016/j.jnt.2017.08.037
![]() |
[15] |
C. Xu, Multiple zeta values and Euler sums, J. Number Theory, 177 (2017), 443–478. https://doi.org/10.1016/j.jnt.2017.01.018 doi: 10.1016/j.jnt.2017.01.018
![]() |
[16] |
C. Xu, Y. Yan, Z. Shi, Euler sums and integrals of polylogarithm functions, J. Number Theory, 165 (2016), 84–108. https://doi.org/10.1016/j.jnt.2016.01.025 doi: 10.1016/j.jnt.2016.01.025
![]() |
[17] |
D. Zheng, Further summation formulae related to generalized harmonic numbers, J. Math. Anal. Appl., 335 (2007), 692–706. https://doi.org/10.1016/j.jmaa.2007.02.002 doi: 10.1016/j.jmaa.2007.02.002
![]() |
[18] |
W. Wang, Y. Chen, Explicit formulas of sums involving harmonic numbers and Stirling numbers, J. Differ. Equ. Appl., 26 (2020), 1369–1397. https://doi.org/10.1080/10236198.2020.1842384 doi: 10.1080/10236198.2020.1842384
![]() |
[19] |
C. Xu, W. Wang, Dirichlet type extensions of Euler sums, C. R. Math., 361 (2023), 979–1010. https://doi.org/10.5802/crmath.453 doi: 10.5802/crmath.453
![]() |
[20] |
A. Dil, K. N. Boyadzhiev, Euler sums of hyperharmonic numbers, J. Number Theory, 147 (2015), 490–498. https://doi.org/10.1016/j.jnt.2014.07.018 doi: 10.1016/j.jnt.2014.07.018
![]() |
[21] |
I. Mező, A. Dil, Euler-Seidel method for certain combinatorial numbers and a new characterization of Fibonacci sequence, Open Math., 7 (2009), 310–321. https://doi.org/10.2478/s11533-009-0008-5 doi: 10.2478/s11533-009-0008-5
![]() |
[22] |
M. E. Hoffman, Harmonic-number summation identities, symmetric functions, and multiple zeta values, Ramanujan J., 42 (2017), 501–526. https://doi.org/10.1007/s11139-015-9750-4 doi: 10.1007/s11139-015-9750-4
![]() |
[23] |
M. Kuba, A. Panholzer, A note on harmonic number identities, Stirling series and multiple zeta values, Int. J. Number Theory, 15 (2019), 1323–1348. https://doi.org/10.1142/S179304211950074X doi: 10.1142/S179304211950074X
![]() |
[24] | Q. Ma, W. Wang, Infinite series identities on r-Stirling numbers, Rocky Mountain J. Math., in press. |
[25] | J. M. Campbell, K. Chen, An integration technique for evaluating quadratic harmonic sums, Aust. J. Math. Anal. Appl., 18 (2021), 15–19. |
[26] | X. Chen, W. Wang, Apéry-type series via colored multiple zeta values and Fourier-Legendre series expansions, submitted for publication, 2024. |
[27] |
H. Liu, W. Wang, Gauss's theorem and harmonic number summation formulae with certain mathematical constants, J. Differ. Equ. Appl., 25 (2019), 313–330. https://doi.org/10.1080/10236198.2019.1572127 doi: 10.1080/10236198.2019.1572127
![]() |
[28] |
H. Liu, W. Zhou, S. Ding, Generalized harmonic number summation formulae via hypergeometric series and digamma functions, J. Differ. Equ. Appl., 23 (2017), 1204–1218. https://doi.org/10.1080/10236198.2017.1318861 doi: 10.1080/10236198.2017.1318861
![]() |
[29] |
J. M. Campbell, P. Levrie, C. Xu, J. Zhao, On a problem involving the squares of odd harmonic numbers, Ramanujan J., 63 (2024), 387–408. https://doi.org/10.1007/s11139-023-00765-7 doi: 10.1007/s11139-023-00765-7
![]() |
[30] |
A.S. Nimbran, P. Levrie, A. Sofo, Harmonic-binomial Euler-like sums via expansions of (arcsinx)p, RACSAM, 116 (2022), 23. https://doi.org/10.1007/s13398-021-01156-7 doi: 10.1007/s13398-021-01156-7
![]() |
[31] |
W. Wang, C. Xu, Alternating multiple zeta values, and explicit formulas of some Euler-Apéry-type series, Eur. J. Combin., 93 (2021), 103283. https://doi.org/10.1016/j.ejc.2020.103283 doi: 10.1016/j.ejc.2020.103283
![]() |
[32] |
X. Wang, X. Yuan, Infinite summation formulas involving the Riemann-zeta function, Integers, 20 (2020), 1–15. https://doi.org/10.5281/zenodo.10776560 doi: 10.5281/zenodo.10776560
![]() |
[33] |
C. Wei, C. Xu, On some conjectural series containing harmonic numbers of 3-order, J. Math. Anal. Appl., 546 (2025), 129306. https://doi.org/10.1016/j.jmaa.2025.129306 doi: 10.1016/j.jmaa.2025.129306
![]() |
[34] | J. Edwards, Differential calculus for beginners, Macmillan, 1893. |
[35] | I. J. Schwatt, An introduction to the operations with series, 2 Eds., New York: Chelsea Publishing Co., 1962. |
[36] |
J. M. Borwein, M. Chamberland, Integer powers of arcsin, Int. J. Math. Math. Sci., 2017 (2007), 19381. https://doi.org/10.1155/2007/19381 doi: 10.1155/2007/19381
![]() |
[37] | J. Lin, W. Wang, Infinite series identities on central binomial coefficients, submitted for publication, 2024. |
[38] |
D. Orr, Generalized log-sine integrals and Bell polynomials, J. Comput. Appl. Math., 347 (2019), 330–342. https://doi.org/10.1016/j.cam.2018.08.026 doi: 10.1016/j.cam.2018.08.026
![]() |
[39] |
H. W. Gould, Announcement, Fibonacci Quart., 10 (1972), 662. https://doi.org/10.1080/00150517.1972.12430893 doi: 10.1080/00150517.1972.12430893
![]() |
[40] |
C. Lupu, Another look at Zagier's formula for multiple zeta values involving Hoffman elements, Math. Z., 301 (2022), 3127–3140. https://doi.org/10.1007/s00209-022-02990-0 doi: 10.1007/s00209-022-02990-0
![]() |
[41] | A. Sofo, Harmonic number sums in higher powers, J. Math. Anal., 2 (2011), 15–22. |
[42] |
J. Zhao, Sum formula of multiple Hurwitz-zeta values, Forum Math., 27 (2015), 929–936. https://doi.org/10.1515/forum-2012-0144 doi: 10.1515/forum-2012-0144
![]() |
[43] |
D. Orr, Generalized rational zeta series for ζ(2n) and ζ(2n+1), Integr. Transf. Spec. Funct., 28 (2017), 966–987. https://doi.org/10.1080/10652469.2017.1393809 doi: 10.1080/10652469.2017.1393809
![]() |