PPNCKH
PPNCKH
PPNCKH
l ∈ {2, 3, 4, 5}
Abstract
For any positive integer l, let Bl (n) denotes the number of l-regular partition
triples of a positive integer n. By employing q-series identities, we prove infinite
family of arithmetic identities and congruences modulo 4 for B2 (n), modulo 2
and 9 for B3 (n),modulo 2 for B4 (n) and modulo 2 and 5 for B5 (n).
Keywords: l-regular partition · Partition triples · Partition congruence ·
q-series identities · Ramanujan’s theta-functions.
Mathematics Subject Classification (2010) 11P83 · 05A15 · 05A17
1. Introduction
where
∞
Y
(a; q)∞ = (1 − aq n ). (2)
n=0
∗ Correspondingauthor
Email addresses: nipennak@yahoo.com (Nipen Saikia), cboruah123@gmail.com
(Chayanika Boruah)
Preprint submitted to Theory of Probability and its Applications April 18, 2024
The arithmetic properties of l-regular partitions have been studied by many
authors, for example see [2, 3, 6, 7, 8, 9, 10, 11, 14, 15, 16] and references there
in.
A partition triple (λ1 ,λ2 ,...,λk ) of a positive integer n is called l-regular
partition triple if none of λi , i = 1, 2, and 3, is divisible by l. If Bl (n) denotes
the the number of l-regular partition triples of a positive integer n then Bl (n)
is defined as
∞
X (q l ; q l )3∞
Bl q n = , (3)
n=0
(q; q)3∞
2. Preliminaries
Ramanujan’s general theta-function f (a, b) is defined by
∞
X
f (a, b) = an(n+1)/2 bn(n−1)/2 , |ab| < 1. (4)
n=−∞
and
∞
X (q 2 ; q 2 )2∞
ψ(q) := f (q, q 3 ) = q n(n+1)/2 = . (6)
n=0
(q; q)∞
2
Proof. Follows easily from binomial theorem.
Lemma 2.2 ([1], Lemma 1.4). For any prime p, we have
2
(q; q)p∞ ≡ (q p ; q p )p∞ (mod p2 ).
Lemma 2.3 ([17], p. 382, Eqn. (3.38)). We have
(q 3 ; q 3 )∞ (q 4 ; q 4 )6 (q 6 ; q 6 )3 (q 4 ; q 4 )2∞ (q 6 ; q 6 )∞ (q 12 ; q 12 )2∞
3
= 2 2 9 ∞ 1 1 ∞2 + 3q . (7)
(q; q)∞ (q ; q )∞ (q 2; q 2)∞ (q 2 ; q 2 )7∞
Lemma 2.4 ([7], Theorem 2.2). For any prime p ≥ 5, we have
p−1
2
X 2 3p2 +(6k+1)p 3p2 −(6k+1)p ±p−1 p2 −1 p2
f (−q) := (q; q)∞ = (−1)k q (3k +k)/2
f (−q 2 ,−q( 2 )
) + (−1) 6 q 24 f (−q )
,
k=− p−1
2
(8)
(
p−1
where ±p−1
:= 6 , p ≡ 1 (mod 6)
6 −p−1
6 , p ≡ −1 (mod 6)
±p−1 3k2 +k p2 −1
Furthermore, if − p−1 p−1
2 ≤ k ≤ 2 and k ̸= 2 , then 2 ̸≡ 24 (mod p).
Lemma 2.5 ([7], Theorem 2.1). For any odd prime p, we have
p−3
2
X k2 +k p2 +(2k+1)p p2 −(2k+1)p p2 −1 2
ψ(q) = q 2 f (q 2 ,q 2 )+q 8 ψ(q p ), (9)
k=0
where
k2 + k p2 − 1 p−3
̸≡ (mod p), 0 ≤ k ≤ .
2 8 2
Lemma 2.6 ([5], p. 267, Eqn. (11.6)). We have
(q; q)6∞
= F −5 (q) − 11q − q 2 F 5 (q), (10)
(q 5 ; q 5 )6∞
1
where F (q) := q − 5 (q) and R(q) is the Rogers-Ramanujan continued fraction
defined by
1
q5 q q2 q3
R(q) := + + + + ..., |q| < 1
1 1 1 1
Lemma 2.7 ([4], p. 648, Eqn. (2.2)). For any integer k ≥ 1, we have
32k − 1
p6 (32k n + ) = 32k p6 , (11)
4
where
∞
X
p6 q n = (q; q)6∞ . (12)
n=0
3
Lemma 2.8 ([9], Theorem 1). We have
(q 5 ; q 5 )∞ (q 8 ; q 8 )∞ (q 20 ; q 20 )2 (q 4 ; q 4 )3 (q 10 ; q 10 )∞ (q 40 ; q 40 )∞
= 2 2 2 40 40 ∞ + q 2 2 ∞3 8 8 .
(q; q)∞ (q ; q )∞ (q ; q )∞ (q ; q )∞ (q ; q )∞ (q 20 ; q 20 )∞
which is the α = 0 case of (13). Suppose that (13) holds for α ≥ 0. Employing
Lemma 2.4 in (15), we obtain
∞
X (p2α − 1) n
B2 (p2α n + )q (16)
n=0
8
∞
X p2α+2 − 1 n
B2 (p2α+1 n + )q ≡ (q p ; q p )∞ (q 2p ; q 2p )∞ (mod 4). (19)
n=0
8
4
Again, extracting terms involving q pn from both sides of (19) and replacing
p
q by q, we obtain
∞
X p2(α+1) − 1 n
B2 p2(α+1) n + q ≡ (q; q)∞ (q 2 ; q 2 )∞ (mod 4), (20)
n=0
8
p2(α+1)−1
B2 (p2α+1 (pn + j) + )≡0 (mod 4). (21)
8
Proof. Equating the coefficients of q pn+j , for 1 ≤ j ≤ p − 1, from both sides of
(19), we arrive at the desired result.
Theorem 4.1. For any odd prime p and any non-negative integers α and n,
we have
∞
X p2α − 1 n
B3 (2p2α n + )q ≡ ψ(q) (mod 2). (22)
n=0
4
∞
X (q 3 ; q 3 )3∞
B3 q n = (23)
n=0
(q; q)3∞
∞
X (q 3 ; q 3 )∞
B3 q n ≡ (q 6 ; q 6 )∞ (mod 2). (24)
n=0
(q; q)3∞
∞
X
B3 (2n)q n ≡ (q; q)3∞ (mod 2). (25)
n=0
5
Using Lemma 2.1 in (6), we find that
∞
X
B3 (2n)q n ≡ ψ(q) (mod 2), (27)
n=0
which is the α = 0 case of (22). Now, suppose (22) is true for some α ≥ 0.
Employing Lemma 2.5 in (22), we obtain
p−3
∞
X p2α − 1 n X 2
k2 +k p2 +(2k+1)p p2 −(2k+1)p 2 2
2α
B3 (2p n + )q ≡ q 2 f (q 2 ,q 2 ) + q (p −1)/8 ψ(q p ).
n=0
4
k=0
(28)
2
Extracting terms involving q pn+(p −1)/8 from both sides of (28), dividing by
2
(p −1)/8
q and replacing q p by q, we obtain
∞
X p2(α+1) − 1 n
B3 (2p2α+1 n + )q ≡ ψ(q p ) (mod 2). (29)
n=0
4
Again, extracting terms involving q pn from both sides of (29) and replacing
p
q by q, we obtain
∞
X p2(α+1) − 1 n
B3 (2p2α+1 n + )q ≡ ψ(q) (mod 2). (30)
n=0
4
p2(α+1) − 1
B3 (2p2α+1 (pn + j) + )≡0 (mod 2). (31)
4
Proof. Equating the coefficients of q pn+j , for 1 ≤ j ≤ p − 1, from both sides of
(29), we arrive at the desired result.
Theorem 4.3. For any integer k ≥ 1, we have
32k − 1
B3 (32k n + )≡0 (mod 9). (32)
4
6
Proof. Using Lemma 2.2 in (23), we obtain
∞
X
B3 q n ≡ (q; q)6∞ (mod 9). (33)
n=0
∞
X 9(p2α − 1) n
B4 (p2α n + )q ≡ (q; q)∞ (q 8 ; q 8 )∞ (mod 2). (35)
n=0
24
∞
X (q 4 ; q 4 )3∞
B4 q n = , (36)
n=0
(q; q)3∞
∞
X
B4 q n =≡ (q; q)∞ (q 8 ; q 8 )∞ (mod 2), (37)
n=0
which is the α = 0 case of (35). Now, suppose that (35) is true for some
α ≥ 0. Employing Lemma 2.4 in (35), we obtain
∞
X 9(p2α − 1) n
B4 (p2α n + )q (38)
n=0
24
3k 2 + k 3m2 + m 9p2 − 9
+ 8. ≡ (mod p), (39)
2 2 24
where −(p − 1)/2 ≤ k, m ≤ (p − 1)/2. The congruence (39) is equivalent to
7
(6k + 1)2 + 8(6m + 1)2 ≡ 0 (mod p). (40)
±p−1
For ( −8
p ) = −1, the congruence (40) holds if and only if m = k = 6 .
2
Extracting the terms containing q pn+(9(p −1))/24 from both side of (39), dividing
2
by q (9(p −1))/24 and replacing q p by q, we obtain
∞
X 9(p2α+2 − 1) n
B4 (p2α+1 n + )q ≡ (q p ; q p )∞ (q 8 p; q 8 p)∞ (mod 2). (41)
n=0
24
Again, extracting terms involving q pn from both sides of (41) and replacing
p
q by q, we obtain
∞
X 9(p2(α+1) − 1) n
B4 (p2(α+1) n + )q ≡ (q; q)∞ (q 8 ; q 8 )∞ (mod 2). (42)
n=0
24
9(p2(α+1)−1
B4 (p2α+1 (pn + j) + )≡0 (mod 2). (43)
24
Proof. Equating the coefficients of q pn+j , for 1 ≤ j ≤ p − 1, from both sides of
(41), we arrive at the desired result.
Theorem 6.1. For any odd prime p and any non-negative integers α and n,
we have
∞
X p2α − 1 n
B5 (4p2α n + )q ≡ ψ(q) (mod 2). (44)
n=0
2
∞
X (q 5 ; q 5 )3∞
B5 q n = , (45)
n=0
(q; q)3∞
8
∞
X (q 8 ; q 8 )∞ (q 20 ; q 20 )2∞ (q 4 ; q 4 )3 (q 10 ; q 10 )∞ (q 40 ; q 40 )∞ 3
B5 q n = ( 2 2 2 40 40
+ q 2 2 ∞3 8 8 ) (46)
n=0
(q ; q )∞ (q ; q )∞ (q ; q )∞ (q ; q )∞ (q 20 ; q 20 )∞
Expanding the right hand side of (46), extracting the terms involving q 2n
and replacing q 2 by q, we obtain
∞
X (q 4 ; q 4 )3∞ (q 10 ; q 10 )6∞ (q 2 ; q 2 )6∞ (q 5 ; q 5 )2∞ (q 20 ; q 20 )∞
B5 (2n)q n = + 3q . (47)
n=0
(q; q)6∞ (q 20 ; q 20 )3∞ (q; q)8∞ (q 4 ; q 4 )∞
∞
X
B5 (2n)q n ≡ (q 2 ; q 2 )3∞ + q(q 10 ; q 10 )3∞ (mod 2). (48)
n=0
∞
X
B5 (4n)q n ≡ (q; q)3∞ ≡ ψ(q) (mod 2), (49)
n=0
where we used (26) in the last step. This shows that (44) is true for the case
α = 0.
Now, suppose (44) is true for some α ≥ 0. . Employing Lemma 2.5 in (44),
we obtain
p−3
∞
X p2α − 1 n X 2
k2 +k p2 +(2k+1)p p2 −(2k+1)p 2 2
2α
B5 (4p n + )q ≡ q 2 f (q 2 ,q 2 ) + q (p −1)/8 ψ(q p )
n=0
2
k=0
(50)
2
Extracting terms involving q pn+(p −1)/8 from both sides of (50), dividing by
(p2 −1)/8
q and replacing q p by q, we obtain
∞
X p2(α+1) − 1 n
B5 (4p2α+1 n + )q ≡ ψ(q p ) (mod 2). (51)
n=0
2
Again, extracting terms involving q pn from both sides of (51) and replacing
q p by q, we obtain
∞
X p2(α+1) − 1 n
B5 (2p2(α+1) n + )q ≡ ψ(q p ) (mod 2). (52)
n=0
2
9
Theorem 6.2. For any odd prime p and integers α ≥ 0 and 1 ≤ j ≤ p − 1, we
have
p2(α+1) − 1
B5 (4p2α+1 (pn + j) + )≡0 (mod 2). (53)
2
Proof. Equating the coefficients of q pn+j , for 1 ≤ j ≤ p − 1, from both sides of
(51), we arrive at the desired result.
Theorem 6.3. For any odd prime p p and any non-negative integers α and n,
we have
∞
X 5p2α − 1 n
B5 (20p2α n + )q ≡ ψ(q) (mod 2). (54)
n=0
2
Proof. Extracting the terms invloving q 10n+1 from (48), dividing by q, and
replacing q 10 by q, we obtain
∞
X
B5 (20n + 2)q n ≡ (q; q)3∞ ≡ ψ(q) (mod 2), (55)
n=0
p−3
∞
X 5p2α − 1 n X 2
k2 +k p2 +(2k+1)p p2 −(2k+1)p 2 2
2α
B5 (20p n + )q ≡ q 2 f (q 2 ,q 2 ) + q (p −1)/8 ψ(q p ).
n=0
2
k=0
(56)
2
Extracting terms involving q pn+(p −1)/8 from both sides of (56), dividing by
2
(p −1)/8
q and replacing q p by q, we obtain
∞
X 5p2(α+1) − 1 n
B5 (20p2α+1 n + )q ≡ ψ(q p ) (mod 2). (57)
n=0
2
Again, extracting terms involving q pn from both sides of (57) and replacing
p
q by q, we obtain
∞
X 5p2(α+1) − 1 n
B5 (20p2(α+1) n + )q ≡ ψ(q) (mod 2). (58)
n=0
2
10
Theorem 6.4. For any odd prime p and integers α ≥ 0 and 1 ≤ j ≤ p − 1, we
have
5p2(α+1) − 1 n
B5 (20p2α+1 (pn + j) + )q ≡ 0 (mod 2). (59)
2
Proof. Equating the coefficients of q pn+j , for 1 ≤ j ≤ p − 1, from both sides of
(57), we arrive at the desired result.
Theorem 6.5. For any positive integers n and α ≥ 0, we have
5α − 1
B5 (5α n + ) ≡ (−1)α B5 (n) (mod 5). (60)
2
Proof. Setting l = 5 n (3), we obtain
∞
X (q 5 ; q 5 )3∞
B5 q n = . (61)
n=0
(q; q)3∞
∞
X
−10
B5 q n ≡ (q 5 ; q 5 )12
∞ (F (q) + 119q 2 + q 4 F 10 (q) − 22qF −5 (q) + 22q 3 F 5 (q)) (mod 5).
n=0
(63)
∞
X
B5 (5n + 2)q n ≡ (−1)(q; q)12
∞ (mod 5), (64)
n=0
11
References
[1] Baruah, N.D., Ahmed, Z.: Congruences modulo p2 and p3 for k dots
bracelet partitons with k = mps .J. Number Theory 151(5), 129–146 (2015).
[2] Baruah, N.D., Ahmed, Z.: New congruences for l-regular partitions for
l{5, 6, 7, 49}. Ramanujan J. doi: 10.1007/s11139-015-9752-2 (2016)
[3] Baruah, N.D., Das, K.: Parity results for 7-regular and 23-regular parti-
tions. Int. J. Number Theory II, 2221–2238 (2015).
[4] Baruah, N.D., Sarmah, B.K.: Identities and congruences for the general
partition and Ramanujan’s Tau functions. Indian J. Pure Appl. Math.
44(5), 643–671 (2013).
[5] Berndt, B.C.: Ramanujan’s notebooks part, vol. III. Springer, New York
(1991).
[6] Carlson, R., Webb, J.J.: Infinite families of congruences for k -regular
partitions. Ramanujan J. 33, 329– 337 (2014).
[7] Cui, S.P., Gu, N.S.S.: Arithmetic properties of l - regular partitions. Adv.
Appl. Math. 51, 507–523 (2013).
[8] Dandurand, B., Penniston, D.: l-divisibility of l-regular partition functions.
Ramanujan J. 19, 63–70 (2009).
[9] Hirschhorn, M.D., Sellers, J.A.: Elementary proofs of parity results for
5-regular partitions. Bull. Aust. Math. Soc. 81, 58–63 (2010).
[10] Lin, B.L.S.: An infinite family of congruences modulo 3 for 13-regular
bipartitions. Ramanujan J. doi: 10.1007/s11139-014-9610-7 (2014).
[11] Penniston, D.: Arithmetic of l-regular partition functions. Int. J. Number
Theory 4, 295–302 (2008).
[12] Wang, L.: Arithmetic properties of overpartition triples arXiv:1410.7898v2
[math.NT] 12 (2015).
[13] Wang, L.: Arithmetic identities and congruences for partition triples with
3-cores. Int. J Number Theory. doi:10.1142/S1793042116500627 (2015).
[14] Webb, J.J.: Arithmetic of the 13-regular partition function modulo 3. Ra-
manujan J. 25, 49–56 (2011).
[15] Xia, E.X.W., Yao, O.X.M.: Parity results for 9-regular partitions. Ramanu-
jan J. 34, 109–117 (2014).
[16] Xia, E.X.W., Yao, O.X.M.: A proof of Keith’s conjecture for 9-regular
partitions modulo 3. Int. J. Number Theory 10, 669–674 (2014).
[17] Xia, E.X.W., Yao, O.X.M.: Analogues of Ramanujan’s partition identities.
Ramanujan J. 31, 373–396 (2013).
12