quantum reports
Article
Quantum Computing, Seifert Surfaces,
and Singular Fibers
Michel Planat 1,*
, Raymond Aschheim 2
, Marcelo M. Amaral 2 and Klee Irwin 2
1
Institut FEMTO-ST CNRS UMR 6174, Université de Bourgogne/Franche-Comté,
15 B Avenue des Montboucons, F-25044 Besançon, France
2 Quantum Gravity Research, Los Angeles, CA 90290, USA; raymond@QuantumGravityResearch.org (R.A.);
Marcelo@quantumgravityresearch.org (M.M.A.); Klee@quantumgravityresearch.org (K.I.)
* Correspondence: michel.planat@femto-st.fr
Received: 11 March 2019; Accepted: 17 April 2019; Published: 24 April 2019






Abstract: The fundamental group π1(L) of a knot or link L may be used to generate magic states
appropriate for performing universal quantum computation and simultaneously for retrieving
complete information about the processed quantum states. In this paper, one defines braids whose
closure is the L of such a quantum computer model and computes their braid-induced Seifert surfaces
and the corresponding Alexander polynomial. In particular, some d-fold coverings of the trefoil knot,
with d = 3, 4, 6, or 12, define appropriate links L, and the latter two cases connect to the Dynkin
diagrams of E6 and D4, respectively. In this new context, one finds that this correspondence continues
with Kodaira’s classification of elliptic singular fibers. The Seifert fibered toroidal manifold Σ′, at the
boundary of the singular fiberẼ8, allows possible models of quantum computing.
Keywords: quantum computing; Seifert surfaces; singular fibers
1. Introduction
To acquire computational advantage over a classical circuit, a quantum circuit needs a
non-stabilizer quantum operation for preparing a non-Pauli eigenstate, often called a magic state.
The work about qubit magic state distillation [1] was generalized to qudits [2] and multi-qubits
(see [3] for a review). Thanks to these methods, universal quantum computation (UQC), the ability to
prepare every quantum gate, is possible. A new approach of UQC, based on permutation gates and
simultaneously minimal informationally-complete positive operator-valued measures (MICs), was
worked out in [4,5]. It is notable that the structure of the projective special linear group (or modular
group) Γ is sufficient for getting most permutation-based magic states [6] used by us for UQC and that
this can be thought of in terms of the complement of the trefoil knot in the three-sphere S3 [7].
Let us recall the context of our work compared to the existing literature. Bravyi and Kitaev
introduced the principle of “magic state distillation” [1]: universal quantum computation,
the possibility of getting an arbitrary quantum gate, may be realized thanks to stabilizer operations
(Clifford group unitaries, preparations, and measurements) and an appropriate single qubit
non-stabilizer state, called a “magic state”. Then, irrespective of the dimension of the Hilbert space
where the quantum states live, a non-stabilizer pure state was called a magic state [2]. An improvement
of this concept was carried out in [4,5] showing that a magic state could be at the same time a fiducial
state for the construction of a minimal informationally-complete positive operator-valued measure
(MIC) under the action on it of the Pauli group of the corresponding dimension. Thus, UQC in this
view happens to be relevant both to magic states and to MICs. In [4,5], a d-dimensional magic state
was obtained from the permutation group that organizes the cosets of a subgroup H of index d of a
two-generator free group G. This is due to the fact that a permutation may be realized as a permutation
Quantum Rep. 2019, 1, 3; doi:10.3390/quantum1010003
www.mdpi.com/journal/quantumrep
Quantum Rep. 2019, 1, 3
2 of 11
matrix/gate and that mutually-commuting matrices share eigenstates: they are either of the stabilizer
type (as elements of the Pauli group) or of the magic type. It is enough to keep magic states that are
simultaneously fiducial states for an MIC because the other magic states may loose the information
carried during the computation. A catalog of the magic states relevant to UQC and MICs can be
obtained by selecting G as the two-letter representation of the modular group Γ = PSL(2,Z) [6].
The next step, developed in [7], is to relate the choice of the starting group G to a three-dimensional
topology. More precisely, G is taken as the fundamental group π1(S3 \ L) of a three-manifold M3
defined as the complement of a knot or link L in the three-sphere S3. A branched covering of degree d
over the selected M3 has a fundamental group corresponding to a subgroup of index d of π1(M3) and
may be identified as a sub-manifold of M3; the one leading to an MIC is a model of UQC. In the specific
case of Γ, the knot involved is the left-handed trefoil knot T1 = 31, as shown in [6] and [7] (Section 2).
1.1. Motivation of the Work
It is desirable that the UQC approach of [4–7] be formulated in terms of braid theory to allow
a physical implementation. Braids of the anyon type, which are two-dimensional quasiparticles
with world lines creating space-time braids, are currently very popular [8–10]. Close to this view
of topological quantum computation (TQC) based on anyons, we propose a TQC based on a Seifert
surface defined over a link L. The links in question will be those able to generate magic states
appropriate for performing permutation-based UQC. In our previous work [7], we investigated the
trefoil knot as a possible source of d-dimensional UQC models through its subgroups of index d
(corresponding to d-fold coverings of the T1 three-manifold) (see [7] (Table 1)). More precisely, the link
L7n1, corresponding to the congruence subgroup Γ0(2) of the modular group Γ, builds a relevant
qutrit magic state for UQC whose MIC geometry is related to the Hesse configuration. The link
L6a3, corresponding to the congruence subgroup Γ0(3) of Γ, builds a relevant two-qubit magic state
whose MIC geometry is the generalized quadrangle of order two GQ(2, 2), as for the commutation of
two-qubit Pauli operators. Then, the link identified by the software SnapPy as L6n1 (or sometimes
L8n3), corresponding to the congruence subgroup Γ(2) of Γ, defines a 6-ditMIC with a building block
geometry looking like Borromean rings [6] (Figure 4). As shown in Section 2.2 below, none of the two
links L6n1 and L8n3 are correctly associated with the subgroup Γ(2) of Γ, but the link 633 (related to the
Dynkin diagram ofD̃4) is. The possible confusion lies in the fact that all three links share the same link
group π1(L). Finally, the link along the Dynkin diagram of D4 (with the icosahedral symmetry of H3 in
the induced permutations) is associated with a 12-dimensional (two-qubit/qutrit) MIC corresponding
to the congruence subgroup 10A1 of Γ [6] (Table 1).
1.2. Contents of the Work
As announced in the Abstract, we introduce a Seifert surface algorithm for converting the UQC
models based on the aforementioned links into the appropriate braid representation permitted by
Alexander’s theorem [11]. These calculations are described in Section 2. A Seifert surface is an oriented
surface whose boundary is a given link. Of course, it is not unique. In this paper, to generate a Seifert
surface, one makes use of the braid representation of the link. Since it has a skein relation different
from the one obeyed by anyons, this kind of topological quantum computation cannot be anyon-based.
The skein relation in question is in terms of the Alexander polynomial instead of the Jones polynomial.
In Section 3, taking into account the observation that some of our UQC models are related to affine
Coxeter–Dynkin diagrams, we build a class of UQC models starting with affine Dynkin diagrams of
typeD̃4,Ẽ6, andẼ8, which are singular fibers of minimal elliptic surfaces. Along the way, topological
objects such as the three-torus, the Poincaré dodecahedral space [12], as well as the first amphicosm [13]
are encountered. They are the precursors of the four-manifold topology that is currently under active
scrutiny [14–16]. Its possible role in models of UQC is discussed in the Conclusion.
Quantum Rep. 2019, 1, 3
3 of 11
2. Seifert Surfaces and Braids from d-Fold Coverings of the Trefoil Knot Manifold (or of
Hyperbolic Three-Manifolds)
Alexander’s theorem states that every knot or link can be represented as a closed braid [11].
A Seifert surface F of a knot K or a link L is an oriented surface within the three-sphere S3 whose
boundary ∂F coincides with that knot or link. Given a basis { fk} for the first homology group H1(F;Z)
of F, one defines a Seifert matrix V whose (i, j)th entry is the linking number of the component fi and
the positive push-off f+
j of the component f j along a vector field normal to F. Then, an invariant of L
is the (symmetrized) Alexander polynomial [11,15] (Section 2.7):
∆L(t) = t−r/2 det(V − tVT),
(1)
with VT the transpose of V and r the first Betti number of F. By definition, ∆L(t−1) = ∆L(t).
There exists a property of ∆L(t) called a skein relation (the Jones polynomial used for defining
anyons obeys a different skein relation than the Alexander polynomial [9,17], so that the rules for
braiding are also different from those resulting from the Seifert surfaces). If L+, L0, and L− are links in
S3, with projections differing from each other by a single crossing, as in Figure 1b, then:
∆L+(t)− ∆L−(t) = (t1/2 − t−1/2)∆L0(t).
(2)
When L is a knot K, there is a connection of ∆K(t) with a combinatorial invariant ν of the
three-manifold S3K obtained from the zero-surgery along K in S
3 as follows:
ν(S3K) =
∆K(t)
(t1/2 − t−1/2)2
.
(3)
The invariant ν(S3K) is called Milnor (or Reidemeister) torsion [18,19].
The invariant ν
has the ability to distinguish closed manifolds, which are homotopy equivalent while being
non homeomorphic.
Figure 1. (a) The Seifert surface F for a trefoil knot K and (b) the types of crossings for the skein relation
of a link L.
There exists a procedure for producing infinite families of homeomorphic, but non-diffeomorphic
four-manifolds. It consists of applying a so-called Fintushel–Stern surgery along a torus T using
some knot K. The so-obtained non-diffeomorphic four-manifolds may be distinguished thanks to the
Seiberg–Witten invariant in which the Alexander polynomial ∆K(e2T) appears as a factor [16] (Chapter 12).
Quantum Rep. 2019, 1, 3
4 of 11
The Seifert surface can be drawn from the braid representation. A good reference is [20] and a
related website [21]. The software SeifertView provides a visualization of the Seifert surface [22].
In practice, to obtain the braid representation and the corresponding Alexander polynomial,
we proceed as follows. In Figures 2–4 below, the braids are oriented from left to right and thus
provide an orientation to the corresponding links. With the software SnapPy [23], one defines the
link from its name (e.g., M = Manifold (“31”) for the trefoil knot K = T1) or from its PD representation
available after drawing the link in the pink editor (e.g., trefoil = [(6, 4, 1, 3), (4, 2, 5, 1), (2, 6, 3, 5)], L =
Link(trefoil) and L.braid_word() for obtaining the braid associated with T1 as [-1,-1,-1], and L.braid_matrix()
for obtaining the Seifert matrix V). Then, with Magma software [24], the Alexander polynomial follows
(det:= Determinant (u ∗V − v ∗ Transpose(V)); det; to obtain u2 − u ∗ v + v2 or t− 1 + t′ after replacing
u by t1/2 and v by t′ = t1/2).
Results are summarized in Table 1.
Table 1. A few models of universal quantum computation (UQC) [7,25] translated into the language of
braids and their Seifert surfaces. The source is a knot (such as the trefoil knot) or a link, and the target
is a link L associated with a degree d covering of the L-manifold that defines an appropriate magic state
for UQC and a corresponding minimal informationally-complete (MIC) measure. Cases d = 3, 4, 5, ...
correspond to the Hesse configuration, to the generalized quadrangle of order two GQ(2, 2) (also called
a doily), to the Petersen graph. The notation for the braids is that of [20]. The notation t′ means t−1.
Source
Target
MIC
Braid Word
Alexander Polynomial
trefoil
L7n1
QTHesse
(ab)3b
t5/2 − t3/2 + t′(3/2) − t′(5/2)
.
L6a3
2QBdoily
ABCDCbaCdEdCBCDCeb
−3t1/2 + 3t′(1/2)
.
633
6-ditMIC
(ab)3
t2 − t− t′ + t2
.
D4 Dynkin
2QB-QT MIC
ABCCbaCCBCCb
−t3/2 + 3t1/2 + t′(3/2) − 3t′(1/2)
fig. eight
L10n46
2QB doily
abCbabbcBc
−t5/2 + 4t3/2 − 4t1/2 + · · ·
.
L14n55217
7-dit MIC
AbbcbcbDacBacdcb
−t4 + 7t3 − 11t2 + 8t− 6 + · · ·
Whitehead
L12n1741
QT Hesse
AbcDEFeDCBDacBdcdEdfCbdCddddeD
−2t3 + 6t2 − 6t + 4 + · · ·
.
L13n11257
5-dit MIC
AbCCbDaCBcDcDcbCD
t9/2 − 6t7/2 + 15t5/2 − 21t3/2
+21t1/2 + · · ·
623 = L6a1
L12n2181
QT Hesse
ABcdEFceGbdFaedCBcdEdfcEgbdfedc
4t5/2 − 12t3/2 + 16t1/2 + · · ·
.
L14n63905
2QB doily
AbCddEdFedcBdaEdfCbceDccDcBC
t4 − 7t3 + 22t2 − 41t + 50 + · · ·
L6a5
L14n63788
QT Hesse
ABCdEEEFEDcebdacEbEED
t4 − 2t3 + 2t− 2 + · · ·
ceDefedCeBdCEDe
Before going further, let us recall the homomorphism between the conjugacy classes of subgroups
of index d of a group G and the d-fold coverings of a manifold M whose fundamental group is
G = π1(M) [26]. This relationship is not one-to-one (not an isomorphism) in the sense that a π1(M)
may characterize distinct manifolds M. A simple example is the number ηd of d-coverings of a
manifold with characteristic 2g− 2 = 0 (with g the genus) where ηd equals the sum of divisor function
σ(d) [27] (Section 3.4). Similarly, it is found in Section 2.2 that distinct links L =L6n1, L8n3, 633, and LK
define manifolds with a fundamental group of common cardinality sequence ηd(L).
2.1. The Braids Built from the Trefoil Knot that Are Associated with the Qutrit Link L7n1 and the Two-Qubit
Link L6a3
As announced in the Introduction, one refers to [7] (Table 1), which lists the topological properties
of d-fold coverings d = 1 . . . 8 of the trefoil knot manifold, as obtained from SnapPy, and also identifies
the corresponding congruence subgroups of Γ previously investigated in [6]. From now on, one denotes
a, b, . . . the generators of the fundamental group π1, one uses the notation A = a−1, B = b−1,. . . for their
inverses, and (.,.) means the group theoretical commutator of the entries. The link L7n1 corresponds to
the congruence subgroup Γ0(2) of Γ. Its fundamental group π1 =
〈
a, b|(a, B2)
〉
builds a qutrit magic
state for UQC of the type (0, 1,±1) and an MIC with the Hesse geometry [7] (Figure 1a). Figure 2a
Quantum Rep. 2019, 1, 3
5 of 11
is the drawing of L7n1. The braid representation (ab)3b is in Figure 2b, and the Seifert surface is in
Figure 2c.
Figure 2.
(a) The link L7n1 defining the qutrit MIC, (b) the braid representation, and (c) the
corresponding Seifert surface. (d) The link L6a3 defining the two-qubit MIC, (e) the braid representation,
and (f) the Seifert surface.
The link L6a3 corresponds to the congruence subgroup Γ0(3) of Γ. Its fundamental group π1 =
〈
a, b|(a, b3)
〉
builds a two-qubit magic state for UQC of the type (0, 1,−ω6, ω6 − 1), ω6 = exp( 2iπ
6 ),
as well as an MIC with the geometry of the generalized quadrangle of order two GQ(2, 2) (sometimes
called a doily) [7] (Figure 1b). Figure 2d is the drawing of L6a3, Figure 2e the corresponding braid
ABCDCbaCdEdCBCDCeb, and Figure 2f the Seifert surface.
The Alexander polynomials are given in Table 1.
2.2. The Braid Built from the Trefoil Knot that Is Associated with the 6-dit Link 633 and Related Braids with the
Same Fundamental Group
There are eight conjugacy classes of subgroups of index six of the modular group Γ corresponding
to eight six-fold coverings over the trefoil knot manifold. They are listed and identified in [7] (Table 1).
We are first interested in the unique regular covering of degree six with homology group Z+Z+Z
(with Z the group of integers) and three cusps corresponding to the congruence subgroup Γ(2), which
we denoteM6,. The fundamental group is π1 = 〈a, b, c|(a, B), (b, C)〉. The cardinality sequence of
subgroups for the fundamental group of this particular covering is that of the link 633:
ηd(6
3
3) = [1, 7, 16, 60, 122, 794, 4212, 35276, 314949, . . .]
(4)
It turns out that every degree six covering of the trefoil manifold leading to a 6-dit MIC (with a
magic state of the type (0, 1, ω6 − 1, 0,−ω6, 0), ω6 = exp( 2iπ
6 )) shares the same fundamental group.
ForM6, SnapPy randomly provides several choices such as L6n1 or L8n3, which of course, share the
same cardinality sequence as 633. In the Knot Atlas at “http://katlas.org/wiki/L6n1”, one finds the
sentence “L6n1 is 633 in Rolfsen’s table of links”. However, that seems to be a wrong statement.
Quantum Rep. 2019, 1, 3
6 of 11
Let us observe that by performing (−2, 1) surgery on all three cusps ofM6 or of 633 and introducing
the Dynkin diagramẼ6 of affine E6 (see Section 3 for details), one gets:
ηd[M6(−2, 1)] = ηd[633(−2, 1)] = ηd(Ẽ6)
= [1, 1, 4, 2, 1, 6, 3, 2, 10, 1, 1, 19, 3, 3, 14, 3, 1, 36, 3, 2, . . .],
while performing (−2, 1) surgery on cusps of L6n1 and introducing the Dynkin diagram of E6, one gets:
ηd[L6n1(−2, 1)] = ηd(E6) = ηd(2T)
= [1, 0, 1, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, . . .],
where 2T is the binary tetrahedral group.
We do not provide the result of performing (−2, 1) surgery on L8n3, which provides a still
different result that we could not identify. We conclude that the correct identification of the manifold
M should be 633, although we do not yet have a rigorous proof.
The link 633, the corresponding braid (ab)
3, and the Seifert surface are shown in Figure 3a–c.
Switching the up/down positions of circles at two points (as shown in Figure 2c) provides the
Kirby link LK drawn in [28] (Figure 3) that we reproduce in Figure 3d (applying the surgeries as
LK(4, 1)(1, 1)(2, 1) to red, blue, and green circles, one gets the Brieskorn sphere Σ(2, 3, 5), alias the
Poincaré dodecahedral space). The corresponding braid word aBabAb and its Seifert surface are in
Figure 3e,f. It is notable that (−2, 1) surgery on K leads to the Dynkin diagram for A3 of Weyl group
S4. Both links 633 and K have the same Alexander polynomial t
2 − t− t′ + t2.
Figure 3. (a) The link 633 corresponding to the 6-dit MIC and the congruence subgroup Γ(2) of Γ, (b) the
braid representation, and (c) the Seifert surface. (d) The Kirby link LK (see the arrows for the up/down
changes), (e) the braid representation, and (f) the Seifert surface (observe the color changes).
The Six-Cover of the Trefoil Knot Manifold Corresponding to the Congruence Subgroup 3C0 of Γ
Let us conclude this subsection by another observation concerning the six-cover (that we denote
M′) of the trefoil knot manifold identified in [7] (Table 1) corresponding to the congruence subgroup
3C0. Again, the cardinality sequence of subgroups of π1(M′) is that of 633, L6n1, L8n3, or LK, but M
′
can be distinguished from the manifolds corresponding to these links since one gets under −2-surgery
Quantum Rep. 2019, 1, 3
7 of 11
ηd[M′(−2, 1)] = ηd[D̃4], whereD̃4 is the Dynkin diagram of affine D4 (as well as the smallest elliptic
singular fiber of Kodaira’s classification; see Figure 5a). Kodaira’s classification may be found in [16]
(p. 320) or [29] (Table 1).
2.3. The Braid Built from the Trefoil Knot that Is Associated with the Two-Qubit/Qutrit MIC with Icosahedral
Symmetry of the Permutation Representation
In [6] (Table 1), the first author identified a two-qubit/qutrit MIC corresponding to the congruence
subgroup 10A1 of the modular group Γ. The geometry of this MIC is that of the four-partite graph
K(3, 3, 3, 3). More precisely, the subgroup of index 12corresponding to 10A1 is generated as G =
〈a, b, c, d|(a, b), (a, c), (a, t)〉, which corresponds to the fundamental group of a link along the Dynkin
diagram of D4. It builds a magic state for UQC of the type (1, 1, 0, 0, 0, 0,−ω6,−ω6, 0, ω6− 1, 0, ω6− 1).
Under (−2, 1) surgery on all four components of the link attached to D4, one recovers the quaternion
group. The permutation representation that organizes the cosets of 10A1 in Γ is the icosahedral
group Z2 × PSL(2, 5), alias the Coxeter group of the Dynkin diagram H3. The braid word and the
corresponding Alexander polynomial are given in Table 1, and the Seifert surface is in Figure 4.
Figure 4.
(a) The Dynkin diagram for the D4 manifold attached to the 2QB-QT MIC, (b) the
corresponding braid ABCCbaCCBCCb, and (c) the Seifert surface.
2.4. Braids from d-Fold Coverings of Hyperbolic Three-Manifolds
Models of UQC from MICs are also sometimes associated with links, as already recognized
in [7,25]. Some of them are listed in Table 1 together as a corresponding braid word and
Alexander polynomial.
2.4.1. The Hyperbolic Link L10n46 and Its Zero-Surgery
Let us focus on the figure-eight knot three-manifold that we callMK4a1. The low degree coverings
overMK4a1 are three-manifolds that may be identified with SnapPy (see [7] (Table 2)). In particular,
there are two coverings of degree four. One is cyclic with a single cusp and homology group Z/13 +
Z/15 + Z and corresponds to the three-manifold otet0800007. The second one is irregular with two
cusps, has the homology group Z+Z, and corresponds to the complement of the link L10n46 (alias
the hyperbolic three-manifold otet0800002) (there is a mistake in [7] (Table 2). Some items have to
be switched in the two degree four coverings over the figure-eight knot manifold.). The latter case
defines a 2QB MIC and may be realized as a braid. The braid word is abCbabbcBc, and the Alexander
polynomial is −t5/2 + 4t3/2 − 4t1/2 + · · · .
Quantum Rep. 2019, 1, 3
8 of 11
Performing zero-surgery on both cusps of L10n46, one gets the same cardinality sequence, that
from zero-surgery on the single cusp of figure-eight knot K4a1, as follows:
ηd[L10n46(0, 1)] = ηd[K4a1(0, 1)] = [1, 1, 1, 2, 2, 5, 1, 2, 2, 4, 3, 17, 1, 1, 2, . . .].
The manifold defined by zero-surgery on the figure-eight knot is denoted Y in [30] where it is
viewed as the boundary of a four-manifold with the rational homology of S1 × D3 (a circle times a
three-disk). It is found that any homology sphere obtained by integral surgery on Y bounds a rational
homology ball.
The manifold Y is also investigated in [31] as a spontaneous surgery producing a manifold that is
a fiber bundle over S1 with the fiber a torus and an Anosov monodromy. The manifold has a “Sol”-type
geometric structure in Thurston’s classification of three-manifolds.
On our side, one finds that MICs result from irregular coverings of degrees 4, 6, 9 and 11 (this
is emphasized with bold characters in the above sequence). The first case (leading to the 2QB MIC)
has the homology Z and the same cardinality structure of subgroups as Y itself; the permutation
group organizing the cosets is the permutation representation of the congruence subgroup Γ0(3)
of the modular group Γ. In the second case (leading to a two-valued 6-dit MIC), the homology is
Z/4 +Z/4 +Z, and the permutation group organizing the cosets is the permutation representation
of the congruence subgroup 3C0 of Γ. The next case corresponds to a two-valued 2QT MIC (QT
means qutrit).
Finally, from (3), one mentions that with zero-surgery on the figure-eight knot K = K4a1, one gets
the Reidemeister torsion of K, and it is given through the Alexander polynomial ∆(K) = −t + 3− t′.
2.4.2. Further Results
As shown in Table 1, there also exists a seven-cover overMK4a1 that is associated to a 7-dit MIC
and the link L14n55217 (alias the three-manifoldM′ =otet1400002).
Other models happen to be more complicated. We do not describe them in more detail.
3. Quantum Computing from Affine Dynkin Diagrams
In the previous section, we found that some MICs for UQC (our approach of universal quantum
computing with complete quantum information) relate to Coxeter–Dynkin diagrams:Ẽ6 andD̃4 for
the 6-dit MIC and D4 for the two-qubit/qutrit MIC. This is an unexpected observation that we would
like to complete by another one: the possibility of defining UQC from the singular fiber I I∗ =Ẽ8 of
Kodaira’s classification of minimal elliptic surfaces (see Figure 5c). This classification is used in the
understanding of some aspects of the four-manifold topology as shown in [16] (p. 320); see also [32]
for a different perspective.
Figure 5. A few singular fibers in Kodaira’s classification of minimal elliptic surfaces. (a) Fiber I∗0 (alias
D̃4), (b) fiber IV∗ (aliasẼ6), and (c) fiber I I∗ (aliasẼ8).
Quantum Rep. 2019, 1, 3
9 of 11
Taking π1(Ẽ8) as the fundamental group of affine E8, the subgroup structure of π1(Ẽ8) has the
following cardinality list:
ηd(Ẽ8) = [1, 1, 2, 2, 1, 5, 3, 2, 4, 1, 1, 12, 3, 3, 4,
. . .]
(5)
where the bold characters mean that at least one of the subgroups leads to an MIC, as in [7]. It is
worthwhile to observe that the boundary of the manifold associated toẼ8 is the Seifert fibered toroidal
manifold [33], denoted Σ′ in [7] (Table 5). It may also be obtained by zero-surgery on the trefoil knot T1.
For the sequence above, the coverings are:
[Ẽ8,Ẽ6, {D̃4,Ẽ8}, {Ẽ6,Ẽ8},Ẽ8, {BR0,D̃4,Ẽ6}, {Ẽ8}, {Ẽ6}, {D̃4,Ẽ8},Ẽ6,
Ẽ8, {BR0,D̃4,Ẽ6,Ẽ8}, {Ẽ8}, {Ẽ6}, {D̃4,Ẽ8}, · · · ]
One observes that the subgroups/coverings are fundamental groups forẼ8Ẽ6,D̃4, or BR0, where BR0
is the manifold obtained by zero-surgery on all circles of Borromean rings. The cardinality sequence of
subgroups of BR0 is:
ηd(BR0) = [1, 7, 13, 35, 31, 91, 57, 155, 130, 217, · · · ]
(6)
which is recognized as A001001 in Sloane’s encyclopedia of integer sequences with the title “Number
of sublattices of index d in generic three-dimensional lattice”.
As already given in Section 1, the subgroup structure of π1(Ẽ6) has the following cardinality list:
ηd(Ẽ6) = [1, 1, 4, 2, 1, 6, 3, 2, 10, 1, . . .].
(7)
For this sequence, the coverings are:
[Ẽ6,Ẽ6, {BR0,Ẽ6},Ẽ6,Ẽ6, {BR0,Ẽ6}, {Ẽ6}, {Ẽ6}, {BR0,Ẽ6},Ẽ6 · · · ].
The subgroup structure forD̃4 is:
ηd(D̃4) = [1, 7, 5, 23, 7, 39, 9, 65, 18, 61, · · · ]
(8)
which corresponds to A263825 in Sloane’s encyclopedia of integer sequences with the title “Total
number cπ1(B1)(n) of n-coverings over the first amphicosm B1’ [13].
To be exhaustive, let us mention thatẼ7 is I I I∗ Kodaira’s singular fiber. Following [29] (Table 1),
it can be obtained by (−2, 1)-surgery on the link L4a1 (in [29] (Table 1), I I∗ is 31(0, 1), I I I∗ is
L4a1(−2, 1), and IV∗ = 633(−2, 1), as one expects). Observe that L4a1 has the same fundamental
group as L7n1 (see Section 2.1) and ηd[L4a1(−2, 1)] = ηd(Ẽ7) = [1, 3, 1, 7, 3, 5, 1, 16, 2, 11, · · · ].
Ẽ7 has coverings of type BR0,D̃4, andẼ7, which we do not detail here.
Reidemeister Torsion of the Manifold Σ′
As a final note for this section, according to (3), the “twisted” Reidemeister torsion νt of the
three-manifold obtained from zero-surgery along a knot K in S3 is the Alexander polynomial of K.
The zero-surgery on the trefoil knot T1 = 31 defines the Seifert fibered toroidal manifold Σ′ introduced
in Section 3, and one gets νt(S3T1) = νt(Σ
′) = t− 1 + t′.
Let BR = L6a4 be the Borromean rings and the manifold BR0 as above (obtained by zero-surgery
on all circles of BR) and BR1 be the manifold obtained by zero-surgery on two circles of BR.
The cardinality sequence ηd(BR0) is as in (6), and ηd(BR1) is found as in (4). In principle, one
can compute the Reidemeister torsion for BR0 and BR1 [19] (Section 2.4). This is left open in this paper.
Quantum Rep. 2019, 1, 3
10 of 11
4. Conclusions
In the first part of this paper, it has been found that some coverings of the trefoil knot, or of
other knots or links, that are used for informationally-complete universal quantum computing are
three-manifolds originating from a link within the three-sphere. From the braid representation of such
links, we defined Seifert surfaces and computed the Alexander polynomials. We pointed out that
some coverings of the trefoil knot (of index 6 and index 12, respectively) connect to the symmetry
of Dynkin diagrams (affine E6 and D4). In the second part, it has been found that the singular fiber
Ẽ8 may be used to generate UQC and that its coverings are the singular fibersẼ6 andD̃4 (alias the
first amphicosm), as well as the manifold BR0 obtained by zero-surgery on all circles of Borromean
rings. Generalizing the analysis to coverings ofD̃4 and BR0, one finds that the three-torus enters the
game [12].
While a Jones polynomial controls an anyon braid, the Alexander polynomial controls the
braid-induced Seifert surface. Further work is in progress to connect our work to the theory of
four-manifolds and the Seiberg–Witten invariant [14,16]. For instance, braided surfaces wrapped
around singularities within four-manifolds support quasiparticles [34]. Another motivation for
three- and four-dimensional topological quantum computing concerns the foundations of quantum
gravity [35]. Topological invariants are important in topological quantum field theory and its
application to quantization in general relativity, through loop quantum gravity [36,37]. In this context,
starting from a pure topological action, it is possible to address the quantization of general relativity
with its diffeomorphism invariance. Topological invariants play an essential role in defining and
computing quantum amplitudes in these models, and it is tentative to associate them with topological
quantum computing.
Author Contributions: All authors contributed substantially to the research. Conceptualization: M.P. and K.I.;
methodology, M.P.; software, M.P. and R.A.; validation, R.A. and M.M.A.; formal analysis, M.P. and M.M.A.;
investigation, M.P., R.A., and M.A.; resources, R.A and K.I.; data curation, M.P.; writing, original draft preparation,
M.P.; writing, review and editing, M.P. and M.M.A.; supervision, K.I.; project administration, M.P. and K.I.;
funding acquisition, K.I.
Funding: The research was funded by Quantum Gravity Research, Los Angeles, CA.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Bravyi, S.; Kitaev, A. Universal quantum computation with ideal Clifford gates and noisy ancillas. Phys. Rev.
2005, A71, 022316. [CrossRef]
2.
Veitch, V.; Mousavian, S.A.; Gottesman, D.; Emerson, J. The resource theory of stabilizer quantum
computation. New J. Phys. 2014, 16, 013009. [CrossRef]
3.
Seddon, J.R.; Campbell, E. Quantifying magic for multi-qubit operations. arXiv 2019, arXiv:1901.03322.
4.
Planat, M.; Haq, R.U. The magic of universal quantum computing with permutations. Adv. Math. Phys. 2017,
217, 5287862. [CrossRef]
5.
Planat, M.; Gedik, Z. Magic informationally complete POVMs with permutations. R. Soc. Open Sci. 2017,
4, 170387. [CrossRef]
6.
Planat, M. The Poincaré half-plane for informationally complete POVMs. Entropy 2018, 20, 16. [CrossRef]
7.
Planat, M.; Aschheim, R.; Amaral, M.M.; Irwin, K. Universal quantum computing and three-manifolds,
Universal quantum computing and three-manifolds. Symmetry 2018, 10, 773. [CrossRef]
8.
Freedman, M.; Kitaev, A.; Larsen, M.; Wang, Z. Topological quantum computation. Bull. Am. Math. Soc.
2003, 40, 31. [CrossRef]
9.
Kauffman, L.H.; Lomonacco, S.J. Braiding, Majorana fermions, Fibonacci particles and topological quantum
computing. Quant. Inf. Proc. 2018, 17, 201. [CrossRef]
10. Nayak, C. Non-Abelian Anyons and Topological Quantum Computation. Rev. Mod. Phys. 2008, 80, 1083.
[CrossRef]
Quantum Rep. 2019, 1, 3
11 of 11
11. Adams, C.C. The Knot Book, An Elementary Introduction to the Mathematical Theory of Knots; W. H. Freeman and
Co.: New York, NY, USA, 1994.
12. Weeks, J.R. The Shape of Space, 2nd ed.; Chapman & Hall: Boca Raton, FL, USA, 2001; 408p.
13. Chelnokov, G.; Deryagina, M.; Meldnykh, A. On the coverings of Euclidean manifolds B1 and B2.
Commun. Algebra 2017, 45, 1558. [CrossRef]
14. Gompf, R.E.; Stipsicz, A.I. four-Manifolds and Kirby Calculus; Graduate Studies in Mathematics; American
Mathematical Society: Providence, RI, USA, 1999; p.129.
15. Akbulut, S. four-Manifolds; Oxford Graduate Texts in Mathematics 25; Oxford University Press: Oxford,
UK, 2016.
16.
Scorpian, A. The Wild World of four-Manifolds; American Mathematical Society: Providence, RI, USA, 2005.
17. Kauffman, L.H. State models and the Jones polynomial. Topology 1987, 26, 395–407. [CrossRef]
18. Milnor, J. A duality theorem for Reidemeister torsion. Ann. Math. Sec. Ser. 1962, 76, 137–147. [CrossRef]
19. Nicolaescu, L.I. Notes on the Reidemeister Torsion. Available online: http://www.nd.edu/~lnicolae/
(accessed on 1 February 2019).
20. Van Wijk, J.J.; Cohen, A.M. Visualization of Seifert surfaces. IEEE Trans. Vis. Comput. Graph. 2006, 12, 485–496.
[CrossRef]
21. Collins, J. Seifert Matrix Computations. Available online: www.maths.ed.ac.uk/~v1ranick/julia/index.htm
(accessed on 1 November 2018).
22.
van Wijk, J.J.; Cohen, A.M. SeifertView a Program to Visualize Seifert Surfaces in 3D. Available online:
www.win.tue.nl/~vanwijk/seifertview (accessed on 1 November 2018).
23. Culler, M.; Dunfield, N.M.; Goerner, M.; Weeks, J.R. SnapPy, a Computer Program for Studying the
Geometry and Topology of three-Manifolds. Available online: http://snappy.computop.org (accessed
on 1 November 2018).
24.
Bosma, W.; Cannon, J.J.; Fieker, C.; Steel, A. (Eds.) Handbook of Magma Functions; University of Sydney:
Sydney, Australia, 2017; 5914p.
25.
Planat, M.; Aschheim, R.; Amaral, M.M.; Irwin, K. Quantum computing with Bianchi groups. arXiv 2018,
arXiv:1808.06831.
26. Mednykh, A.D. A new method for counting coverings over manifold with finitely generated fundamental
group. Dokl. Math. 2006, 74, 498–502. [CrossRef]
27.
Liskovets, V.A.; Mednykh, A.D. On the number of connected and disconnected coverings over a manifold.
Ars. Math. Contemp. 2009, 2, 181. [CrossRef]
28. Kirby, R.C.; Scharlemann, M.G. Eight faces of the Poincaré homology three-sphere. In Geometric Topology;
Academic Press: New York, NY, USA, 1979; pp 113–146.
29. Kirby, R.C.; Melvin, P. The E8-manifold, singular fibers and handlebody decompositions. Geom. Topol. Publ.
1998, 2, 233–258.
30. Akbulut, S.; Larson, K. Brieskorn spheres bounding rational balls. Proc. Am. Math. Soc. 2018, 146, 1817–1824.
[CrossRef]
31. Hilden, H.; Lozano, M.T.; Montesinos-Amilibia, J.M. On a remarkable polyhedron geometrizing the figure
eight cone manifolds. J. Math. Sci. Univ. Tokyo 1995, 2 501–561.
32. Aschheim, R.; Irwin, K. Constructing numbers in quantum gravity: Infinions. In Proceedings of the Group32
Conference, Prague, Czech Republic, 9–13 July 2018.
33. Wu, Y.-Q. Seifert fibered surgery on Montesinos knots. arXiv 2012, arXiv:1207.0154.
34. Atiyah, M.; Marcolli, M. Anyons in geometric models of matter. J. High Energ. Phys. 2017, 76. [CrossRef]
35. Asselmeyer-Maluga, T.; Król, J. How to obtain a cosmological constant from small exotic R4. Phys. Dark Univ.
2018, 19 66–77. [CrossRef]
36. Rovelli, C.; Vidotto, F. Covariant Loop Quantum Gravity, 1st ed.; Cambridge University Press: Cambridge,
UK, 2014.
37. Amaral, M.; Aschheim, R.; Irwin, K. Quantum gravity at the fifth root of unity. arXiv 2019, arXiv:1903.10851.
c© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).