Conformal bootstrap analysis for the Yang–Lee edge singularity

Conformal bootstrap analysis for the Yang–Lee edge singularity Prog. Theor. Exp. Phys. 2018, 053I01 (15 pages) DOI: 10.1093/ptep/pty054 Conformal bootstrap analysis for the Yang–Lee edge singularity S. Hikami Mathematical and Theoretical Physics Unit, Okinawa Institute of Science and Technology Graduate University, Okinawa, Onna 904-0495, Japan E-mail: hikami@oist.jp Received January 18, 2018; Revised April 4, 2018; Accepted April 16, 2018; Published May 30, 2018 ................................................................................................................... The Yang–Lee edge singularity is investigated by the determinant method of the conformal field theory. 3 × 3 minors are used for the evaluation of the scale dimension. The agreement with the Padé of  expansion in the region 3 < D < 6 is improved. The critical dimension D , where the scale dimension of scalar  is vanishing, is used for the improvement of the Padé. For the understanding of the intersection point of zero loci of 3 × 3 minors, 2 × 2 minors are investigated in detail; these are connected through Plüker relations. ................................................................................................................... Subject Index I10 1. Introduction The conformal field theory was developed a long time ago [1], and the modern numerical approach was initiated in Ref. [2]. Recent studies using this conformal bootstrap method have obtained some remarkable results for the 3D Ising model [3,4], Yang–Lee edge singularities [5,6], O(N ) models [7–10], and self-avoiding walks [11]. A brief summary of the determinant method for the conformal bootstrap theory is as follows. The conformal bootstrap theory is based on the conformal group O(D,2), and the conformal block G ,L is the eigenfunction of the Casimir differential operator D . The eigenvalue of this Casimir operator is C : D G = C G , 2 ,L 2 ,L C = [( − D) + L(L + D − 2)]. (1) The solutions of the Casimir equation have been studied in Refs. [18–20]. The conformal block G (u, v) has two variables u and v, which denote the cross ratios, u = (x x /x x ) , ,L 12 34 13 24 v = (x x /x x ) , where x = x − x (x is a 2D coordinate). They are expressed as u = zz ¯ and 14 23 13 24 ij i j i v = (1 − z)(1 −¯z). For the particular point z =¯z, the conformal block G (u, v) for the spin-zero ,L (L = 0) case has a simple expression: /2 2 2 z    D D + 1 D z G (u, v)| = F , , − + 1; ,  − + 1; . (2) ,0 z=¯z 3 2 1 − z 2 2 2 2 2 2 4(z − 1) The conformal bootstrap determines  by the condition of the crossing symmetry x ↔ x .For 1 3 practical calculations, the point z =¯z = 1/2 is chosen and, by the recursion relation derived from the Casimir equation of Eq. (1), G (u, v) at z =¯z = 1/2 can be obtained [3]. ,L © The Author 2018. Published by Oxford University Press on behalf of the Physical Society of Japan. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 Funded by SCOAP by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami The conformal bootstrap analysis with a small matrix size has been investigated for the Yang–Lee edge singularity with accurate results for the scale dimensions by Gliozzi and Rago [5,6]. For other models, this bootstrap method for the determinant has been considered [12,15,21,22]. In this paper, we emphasize the importance of the structure of minors along the Plüker relations, as shown in the appendix, and we apply it to the Yang–Lee model for a scale dimension of  . The intersection point of the zero loci of 3 × 3 minors was evaluated and it gave a remarkable value of the critical exponent in Ref. [5]. In these 3 × 3 minors, if the value of the space dimension D is given, the value of  =  is determined from the intersection point. However, if one goes to larger minors, for instance 4 × 4or5 × 5 minors, one need the values of additional scale dimensions of the operator product expansion (OPE). In the study of such large minors, however, the values of  deviate from the  expansion [6]. This discrepancy should be improved by the determinant method or by analysis of the  expansion, since the conformal bootstrap method should be consistent with the  expansion [24,25]. Recently, consistent results were obtained in the O(N ) vector model through the Mellin amplitude [13]. We consider this discrepancy by repeating Gliozzi’s evaluation of the 3 × 3 minors [5] and find that 3 × 3 minors give accurate values that agree with the improved  expansion (see Fig. 8). The nature of the determinant method is still not known, and the convergence to the true value seems to be slow. As we mentioned before, we have to assume the values of the scale dimensions of the OPE, such as spin 4, spin 6, etc., for the large minors. Unfortunately we do not know precisely these higher spin values at present. Therefore, we concentrate on 3 × 3 minors without any assumption of the other scale dimensions of the OPE. The four-point correlation function is given by g(u, v) φ(x )φ (x )φ (x )φ (x ) = (3) 1 2 3 4 2 2 φ φ |x | |x | 12 34 and the amplitude g(u, v) is expanded as the sum of conformal blocks: g(u, v) = 1 + p G (u, v). (4) ,L ,L ,L The crossing symmetry of x ↔ x implies 1 3 φ φ v G (u, v) − u G (v, u) ,L ,L p = 1. (5) ,L φ φ u − v ,L The minor method consists of the derivatives at the symmetric point z =¯z = 1/2 of Eq. (5). √ √ By a change of variables z = (a + b)/2, z ¯ = (a − b)/2, derivatives are taken about a and b. Since the number of equations becomes larger than the number of truncated variables , we need to consider the minors for the determination of the values of . The matrix elements of minors are expressed by φ φ v G (u, v) − u G (v, u) (m,n) ,L ,L m n f = ∂ ∂ | (6) a=1,b=0 ,L a b φ φ u − v and the minors of 2 × 2, 3 × 3, e.g., d , d , are determinants such as ij ijk (m,n) (m,n) d = det f , d = det f , (7) ij ijk ,L ,L 2/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami where i, j, k are numbers chosen differently from (1, ... , 6), following the concise correspondence to (m, n) as 1 → (2, 0),2 → (4, 0),3 → (0, 1),4 → (0, 2),5 → (2, 1), and 6 → (6, 0). In the Yang–Lee case, a fusion rule is simply written as [5] [ ]×[ ]= 1 +[ ]+[D,2]+[ ,4] + ··· . (8) φ φ φ 4 The scalar (L = 0) term in the bootstrap equation of Eq. (6) is denoted as φ φ v G (u, v) − u G (v, u) ,0 ,0 vs0 = . (9) φ φ u − v We use the notation (vsk) (k is a number, related to the degree of the differential, v is a vector, and s is a scalar) for the derivatives of the (vs0) rule in Eq. (6). For L = 2 we use (vtk), and for L = 4we use (vqk). (vxk) is for L = 6. The scalar (L = 0) case is expressed by the hypergeometric function ( =  for the Yang–Lee edge singularity) −1−/2 −/2 3 × 2 2 −/2 vs0 =−2 ys + ys + ys , (10) 2 2 where D  1 D a ys = F , , − + 1; + ,  − + 1: (11) 3 2 2 2 2 2 2 2 2 8(−2 + a) and ys is derivative of ys. We consider the derivative at a point a = 1. For instance, we have 2 × 2 minors: vs1 vs3 vs2 vs3 d = det , d = det . (12) 13 23 vt1 vt3 vt2 vt3 In this Yang–Lee edge singularity, we find that the 2 × 2 minor, d , becomes zero at D = 6, and it provides exact values of  = 2. 2. 2 × 2 minors The intersection points of zero loci of 3 × 3 minors are decomposed to 2 × 2 minors at the critical dimensions D = 6 for the free theory. In the appendix (relation 1), this decomposition is shown. Among several 2 × 2 minors, d , d , d , ... , the most fundamental minors may be d and d , 12 13 23 13 23 which are made of lower derivatives of a and b. Note that, in the Yang–Lee model, the constraint =  is taken, and the 2 × 2 minor analysis corresponds to the 3 × 3 minor analysis of another model such as the Ising model, in which  =  . From the point of the truncation error in the OPE, as discussed in Ref. [15], it is interesting to start from 2 × 2 minors. The Yang–Lee singularity is connected to supersymmetry through the dimensional reduction of a branched polymer, where we have an exact relation of  =  + 1[14,16,17]. We will discuss the dimensional reduction D + 2 → D behavior in the Yang–Lee model in 2 × 2 minors. The 2 × 2 minors consist of two parameters, dimension D and the scaling dimension  , since we have  2 =  ( =  2 ) for the Yang–Lee case, due to the φ theory. φ φ 3/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 1. The intersection of zero loci of 3 × 3 minors (d , d , d , d , d ) for dimension D = 6. The 123 135 134 234 235 axes are (x, y) = ( , Q). The intersection point of the five lines is D = 6,  = 2 with Q = 8, which is φ φ decomposed to the 2 × 2 minor d . The upper fixed point is related to the minor d . 13 23 (i) Free field theory at D = 6 The zero loci of d gives exactly the scale dimension  = 2.0 at D = 6. The minors for vq, 13 φ vx, and vs (with a scalar scale dimension  ) can be zero: vs1 vq1 vs1 vx1 vs1 vs1 det = 0, det = 0, det = 0. (13) vs3 vq3 vs3 vx3 vs3 vs3 For the value  = 2, they yield Q(=  ) = 8,  = 10, and  = 4, respectively. These φ 4 6 ,  ,  ,  are determined exactly by 2 × 2 minors of d . This is due to the free theory, i.e., φ 4 6 13 no interactions between the scale dimensions. It is known that the six dimensions have algebraic identities, which are valid for a free theory in any dimension D [23]. Other minors yield different values; for instance, d gives  = 2.158 at D = 6. These two 23 φ different values of  are understood when we evaluate 3 × 3 minors, as shown in Figs. 1 and 2. There are two fixed points of zero loci of 3 × 3 minors in the six dimensions. The 3 × 3 minors are denoted by d in Eq. (7); for instance, d means ijk 123 ⎛ ⎞ vs1 vt1 vq1 ⎜ ⎟ d = det vs2 vt2 vq2 . (14) 123 ⎝ ⎠ vs3 vt3 vq3 (ii) D = 3 The zero loci of 2 × 2 minor do not provide good results except for D = 6 dimensions. This is due to the situation where the free theory breaks down in general dimensions. We have to consider the loci of 2 × 2 minors of a finite small value due to the interactions, instead of a zero value, for the non-trivial fixed point. The small non-vanishing value of the minor might have an important meaning for the bound of  . The loci of two important 2 × 2 minors d and d are shown φ 13 23 in Fig. 2, in which the value of d is changed by small values from −0.007 to 0.004, and the zero loci of d are also shown in Fig. 3 as a guide. The correct value of  is realized in some 23 φ 4/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 2. The loci of 2 × 2 minors d with values from −0.007 to 0.004 are shown; d = 0 is also shown by a 13 23 green line. The axes are (x, y) = (D,  ). Fig. 3. A close-up of the contour map in Fig. 2 near the  = 0 loci of minor d . The green line shows the φ 13 zero loci of d as a guide. The axes are (x, y) = (D,  ). 23 φ finite-value contours of d for D < 6. Near D = 6, deviation from the zero loci of d is expected 13 13 since  deviates from 2 as the  = 6 − D expansion shows: 2 − 0.5555. φ φ In Fig. 2, the finite-value curves of d converge to a curve starting from  = 0.25 (at D = 1) 13 φ to 2.0 (at D = 4). This curve resembles the correct  contour if the space dimension is shifted from D to D + 2. (The curve is translated to the right in Fig. 2 by two space dimensions by keeping the same value of  .) This dimensional reduction will be discussed in Sect. 4. The finite-value contour map of d shows a valley shape, and one line is located at D = 2 near the exact value  =−0.4, as shown in Fig. 3. In Fig. 3, the value of minor d is 0.0034, which φ 13 gives  =−0.4. The cusp point moves when the value of minor d is changed and its trace φ 13 may give the bound of  . 5/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 4. The zero loci of d[ ,  ] for different dimensions: D = 6 (dark blue), D = 5.5 (red), D = 5 φ  13 (green), D = 4.5 (yellow), and D = 4 (blue). The axes are (x, y) = ( ,  ). In Figs. 2 and 3, contour maps of d with small non-vanishing values are shown. 3. Critical dimension for  = 0 In the Yang–Lee edge singularity, the scale dimension  of φ becomes zero at some dimension between 2 and 3, since it is −0.4 in two dimensions and approximately 0.2 in three dimensions. We call this dimension the critical dimension D , on which the scale dimension  vanishes. In the c φ determinant method, it is known that even small minors give accurate values of the scale dimensions [5,6]. We discuss here small determinants of 2 × 2, 3 × 3, and 4 × 4 matrices for this critical dimension D . Figure 2 shows Z-shaped curves for zero or very small values of the loci of d . This figure can be understood when we plot  versus  for various values of fixed dimensions. The minor of d φ 13 has so far been considered as a function of D and  with the condition of  =  due to the φ  φ Yang–Lee model. However, in general we are able to consider  and  as two free parameters for d , and, for the Yang–Lee case, we put  =  . 13  φ In Fig. 4, we plot the zero loci of d[ ,  ] for different values of D. This minor then becomes φ  13 the usual one including the Ising case. A complicated contour for the zero loci of d is obtained especially around D = 4.4, where it has three solutions for  =  . This solution corresponds to the Z shape around (D, ) = (4, 1.5) in Fig. 2. In addition, we find that the line of  =  of the Yang–Lee condition goes through very near the zero loci of d , but it does not intersect. This corresponds to Fig. 2 with the finite-value loci of d around D = 3. With some value of D, the contours in Fig. 4 (bottom left) and Fig. 5 intersect at the point ( ,  ) = (0, 0). We find that this critical dimension from Fig. 5 is D = 2.6199. (15) As seen in Fig. 3, for a small finite value of d , the contour of d approaches the line  = 0 13 13 φ from both sides, above  > 0 and  < 0. The peak of the contours approaches the critical φ φ 6/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 5. A contour plot of the zero loci of d[ ,  ] for different dimensions: D = 2.6 (blue), D = 2.61 φ  13 (yellow), D = 2.6199 (green). The axes are (x, y) = ( ,  ). dimension D . We consider the derivative of d by the dimension D, and we estimate D as the c 13 c point at which this derivative becomes zero. For  = 0.000 001, we obtain D = 2.605 74. φ c The derivative of d by the space dimension D gives the approximate solution of D in the limit 13 c → 0. The minor d is proportional to  in the limit  → 0, since vs1, vs3 ∼  , and φ 12 φ φ −1 vt1, vt3 ∼  . We consider how the minors become zero in the limit  → 0. For instance, at D = 2.61, we have, for small  , 2 2 2 vs1 = 0.188 , vs2 = 0.139 , vs3 = 0.225 φ φ φ 1.055 0.667 1.078 vt1 = , vt2 = , vt3 = . (16) φ φ φ The coefficients p , p are obtained from φ t vs0 vt0 p 1 = . (17) vs1 vt1 p 0 Since vs0 =− , vt0 = 0.7185/ in the limit  → 0, we obtain a solution of p =−4, p = φ φ φ t a log (a is a constant), which gives the value of the central charge C = 0. With this choice of p = p , the central charge C becomes C =  /p = 0 for this critical dimension D . This means t (D,2) t c that the energy–momentum tensor operator T =  can be neglected since the OPE coefficient (D,2) of this operator p becomes zero. This is the well known c catastrophe, which leads to logarithmic conformal field theory [26]. When neglecting the  term, and only considering vs0 in Eq. (10), (D,2) we obtain the critical dimension D = 2.748 632 ··· . However, this value is too large compared to the expected value and might not be correct. We need further operators to get the correct value. We hope to get the correct value of the critical dimension D by another sophisticated method by taking higher operators. 4 × 4 minors The numbers of zero loci of minors are C = 6!/4!2!= 15. We investigated 6 4 the critical dimensions D by changing D between D = 2.58 and D = 2.59. Remarkably, at 7/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 6. The intersection of zero loci of minors at D = 2.589 53,  = 0. The axes are (x, y) = (Q,  ). D = 2.589 53, all 15 lines intersect at a single point in the contour map of (Q,  ) under the condition  = 0. If we take this as the value of the critical dimension D , we find the following φ c set of results: D = 2.589 53,  = 0, Q = 4.403 97,  = 3.871 27. (18) c φ This critical dimension D obtained with 4 × 4 minors is close to D = 2.6199 from the 2 × 2 c c minor d with the intersection point of  =  = 0. In Fig. 6, all loci coincide but, if we include 13  φ more relevant operators, this value may change. For the  = 0 case, in any dimension D, d can become zero. When  = 0, by the definition φ ij φ of the scale dimension  , the two-point correlation function G(r) = 1/r becomes a constant in the long-range limit r →∞. In the Yang–Lee edge singularity, the exponent of the density σ is related to  as σ = . (19) D − From this relation, σ is vanishing for  = 0. This means that the density is constant at the transition point. Several interesting systems are known in which the density is constant but a phase transition occurs. One example is a localization problem under the random potential. 4. Dimensional reduction The zero loci of the 2 × 2 minor d shows an interesting characteristic linear behavior for D < 4, which is approximated as  = (3D − 2)/5. In Fig. 7, the shift of the zero loci of d φ 13 for D → D + 2 is shown (translation of two dimensions to the right in Fig. 7). The blue line of Fig. 7 for the zero loci of d almost coincides with the red line of the Yang–Lee edge singularity analyzed by the Padé approximation. The red line represents the result of 3 × 3 minors and it can be 8/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 7. The D → D + 2 shifted line of the zero loci of d (yellow line) is shown by the blue line and is consistent with the estimated Yang–Lee line (red) by Padé analysis. The blue and red lines are almost the same for 1 < D < 5.5. approximated as 3D − 8 = . (20) This expression satisfies the exact values of D = 1( =−1), D = 2( =−0.4), and D = 6 φ φ ( = 2). This dimensional shift seems accidental and only appears in a lower truncation, but there is a dimensional reduction that has been proven rigorously. It is known that a branched polymer in D + 2 dimensions is equivalent to a D-dimensional Yang– Lee edge singularity for the critical phenomena. This dimensional reduction has been explained by supersymmetry [28]. A rigorous mathematical proof of this dimensional reduction was shown in Ref. [29]. We will study this dimensional reduction in a separate paper using the conformal bootstrap in a determinant method [14]. The Yang–Lee edge singularity requires  (=  2 ) =  . For a branched polymer due to dimensional reduction there is a relation  =  + 1. This relation is a manifestation of the supersymmetry of the system [16]. Since the blue line is very close to the red line in Fig. 7, one can make an estimation of the critical dimension D , in which  = 0. The intersection of d with the  = 0 line can be evaluated c φ 13 φ very precisely as D = 0.599 547 1444. Adding 2 to this value for the dimensional reduction, it gives the estimation of the critical dimension D as D = 2.599 547 1444, which is very close to other c c estimations of this paper. We discussed this value in the previous section as D = 2.6199 as in Eq. (18). 5. 3 × 3 minors and Padé analysis Up to now, we have mainly discussed 2×2 minors. If we take a spin-4 operator and its scale dimension Q =  , we need to do an analysis of the intersection of the zero loci of the 3 × 3 minors d . 4 ijk 9/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Table 1. Estimated scale dimensions by zero loci of three dimensional minors. Comparison with Pade analysis is shown. The value of parenthesis is estimation from other zero loci nearby. Q  (Padé) φ φ D = 3.0 0.174 343 (0.187 825) 4.341 06 (3.771 24) 0.229 95 D = 3.5 0.499 401 (0.500 969) 5.041 95 (4.375 56) 0.531 53 D = 4.0 0.823 283 (0.815 623) 5.711 52 (4.922 83) 0.831 75 D = 4.5 1.137 55 (1.123 71) 6.333 95 (5.446 45) 1.1300 D = 5.0 1.438 07 (1.419 87) 6.917 16 (5.959 72) 1.4255 D = 5.5 1.724 69 (1.746 82) 7.469 85 (6.466 72) 1.7165 D = 6.0 2.0 8.0 2.0 From the formula of the minors in a 3 × 6 matrix in the appendix, we have a Plüker formula such as (Eq. (A.12)) [126][345]−[123][456]+[124][356]−[125][346]= 0. (21) For instance, at D = 3, we find that the zero loci of d =[123], d =[126], d =[124], and 123 126 124 d =[346] intersect at a point, and the above Plüker formula is satisfied. In Table 1, we present the intersection of three zero loci of minors d , d , and d . The 123 134 124 value of  is obtained from the intersection point of the zero loci of the minors. However, the intersection point is not the only one, and we find at least three different intersection points for each fixed dimension D. In Table 1, we present in parentheses such nearby different intersection points of the three zero loci of the minors. In the  = 6 − D expansion,  is known up to the four-loop level [30]: 2 3 4 = 2 − 0.555 55x − 0.029 4925x + 0.021 845x − 0.039 4773x , (22) where x =  = 6 − D. Including the critical dimension D , for which  is vanishing, the above c φ expansion becomes 2 2 0.55555 = (6 − x − D ) + − x + O(x ) . (23) φ c 6 − D (6 − D ) 6 − D c c c Inserting the value of D = 2.6199 from Sect. 4, it becomes = (3.3801 − x)[0.591 70 + 0.010 695x − 0.005 561 25x 3 4 + 0.004 817 55x − 0.010 2541x + ···]. (24) This expansion is approximated by the [2,2] Padé method, a + a x + a x 0 1 2 = (3.3801 − x) , (25) 1 + b x + b x 1 2 with a = 0.591 70, a = 2.3916, a = 1.0090, b = 4.023 85, b = 1.6419. The curve of this 0 1 2 1 2 Padé is incorporated in Fig. 8. For the values of  of the Yang–Lee singularity in D dimensions, the previous results obtained by Gliozzi [5] are very close to the values of Table 1; for instance, at D = 4,  = 0.823. In another evaluation with more primary operators [6], the values  = 0.8466 in D = 4 and  = 1.455 in φ φ 10/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 8.  is estimated by a [2,2] Padé with D = 2.6199. The dots are the values from Table 1 fora3 × 3 φ c minor analysis of  . D = 5 are obtained, which are larger than the values in Table 1, and a comparison with  expansion shows the apparent deviations, as shown in Fig. 5 of Ref. [6]. Our new analysis of Padé with a fixed value at the critical dimension  = 0 seems more precise, and the  expansion and bootstrap determinant method agree well with each other, as shown in Fig. 8. One of the aims of this paper was to clarify this point by the evaluation of the critical dimension D . Unfortunately, what we have done is a numerical estimation of D by the bootstrap determinant method and it may be an approximation. It is important to obtain an exact value of D , and the comparison with  expansion will be improved with an exact value of D . 6. Summary We have considered theYang–Lee edge singularity in this article by the bootstrap determinant method, initiated by Gliozzi. Although our results for the scale dimension  are same as the 3×3 determinant result of Gliozzi [5], we have improved the consistency with the result of the  expansion. We find that the basic 2 × 2 minor d is important for the determination of the above scale dimensions through the analysis of the intersection points of dimensions one to six, although we used the practical value of  obtained by 3 × 3 minors. We have shown that the value obtained by the intersection of the zero loci of 3 × 3 minors agrees with the result of the  expansion. The discrepancy between the expansion and the determinant method is reduced with the new Padé analysis with the introduction of the critical dimension D . We obtained the critical dimension D from the minor d for the  = 0 and  → 0 limits. The c 13 φ intersection of the zero loci of d[D,  ,  ] at the point  =  = 0 gives the critical dimension φ  13  φ D = 2.6199. We have estimated the critical dimension approximately by other methods, by the peak approaching the finite d (D = 2.6050), the dimensional reduction value (D = 2.5995), and 4 × 4 13 c c minor analysis (D = 2.589). We emphasize that the estimation of the critical dimension D is useful in practical terms for the precise analysis of the Yang–Lee edge singularity between two and six dimensions by the Padé analysis of the  expansion [30]. We have discussed the dimensional reduction property in the Yang–Lee model, using the fact that a branched polymer in D + 2 dimensions is equivalent to a D-dimensional Yang–Lee edge singularity. Acknowledgements The author is grateful to Ferdinando Gliozzi for discussions on the determinant method and the useful suggestion of the critical dimension. He also thanks Edouard Brézin for discussion of the dimensional reduction problem. 11/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami This work is supported by a Japan Society for the Promotion of Science (JSPS) KAKENHI Grant-in-Aid 16K05491. The Mathematica11 system of wolfram.com is acknowledged for this research. Funding Open Access funding: SCOAP . Appendix. Plüker formula Relation 1 (3 × 3 minors) For the determinant d , which is defined as ⎛ ⎞ vs1 vs2 vs3 ⎜ ⎟ d = det vt1 vt2 vt3 ⎝ ⎠ vq1 vq2 vq3 vs1 vs2 vs1 vs3 vs2 vs3 = (vq3)det − (vq2)det + (vq1)det , (A.1) vt1 vt2 vt1 vt3 vt2 vt3 if the first determinant (d ) and second minor (d ) on the right-hand side are zero, then the third 12 13 minor (d ) should be vanishing when d = 0. 23 123 Relation 2 (2 × 4 matrix) The 2 × 4 matrix is denoted as x x x x 11 12 13 14 M = . (A.2) x x x x 21 22 23 24 The Plüker relation is [12][34]−[13][24]+[14][23]= 0, (A.3) where the minor is denoted as [ij]= (x )(x ) − (x )(x ). The application of this formula to minors 1i 2j 1j 2i of our case can be taken as vs1 vs2 vs3 vs4 M = . (A.4) vt1 vt2 vt3 vt4 Thus, in our notation for 2 × 2 minors, d =[12], d =[13], d =[14], d =[23], d =[24], d =[34] (A.5) 12 13 14 23 24 34 and we have an identity d d − d d + d d = 0, (A.6) 12 34 13 24 14 23 which is represented by the tableaux 1 2 1 3 1 4 − + = 0. (A.7) 3 4 2 4 2 3 12/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Note that the third tableau is not ordered by increasing row and column (4 is larger than 3). Another application of this formula is to take the following matrix: vs1 vt1 vq1 vx1 M = . (A.8) vs3 vt3 vq3 vx3 In the D = 6 case, we find that 2 × 2 minors made of the above matrix are vanishing, [12]=[13]= [14]= 0 for  = 2,  = 8, and  = 10. Therefore they satisfy the relation of Eq. (A.3). In this φ 4 6 D = 6 case, we also find that, for such values of  = 2,  = 8,  = 10, the other three sets of φ 4 6 minors are vanishing at D = 6: vq1 vx1 vt1 vx1 [34]= det = 0, [24]= det = 0, vq3 vx3 vt3 vx3 vt1 vq1 [23]= det = 0. (A.9) vt3 vq3 Relation 3 (3 × 6 matrix) (i) The 3 × 6 matrix in the bootstrap minor method is ⎛ ⎞ vs1 vs2 vs3 vs4 vs5 vs6 ⎜ ⎟ vt1 vt2 vt3 vt4 vt5 vt6 . (A.10) ⎝ ⎠ vq1 vq2 vq3 vq4 vq5 vq6 The Plüker relation is obtained by a mutual exchange of numbers: [146][235]+[124][356]−[134][256]+[126][345]−[136][245]+[123][456]= 0 (A.11) This identity is obtained from the first term to the second term by (2 ↔ 6), and from the second to third by (2 ↔ 3), etc. with the sign. We can also use the following relations (obtained similarly): [126][345]−[123][456]+[124][356]−[125][346]= 0 [136][245]+[123][456]+[134][256]−[135][246]= 0; (A.12) from these equations, Eq. (A.11) becomes [146][235]=−3[123][456]−[125][346]+[135][246]. (A.13) In our minor examples, we have d d =−3d d − d d + d d . (A.14) 146 235 123 456 125 346 135 246 The terms on the right-hand side are ordered by increasing row and column. (ii) Another application can be taken as ⎛ ⎞ vs1 vt1 vq1 vx1 vs1 vs1 ⎜ ⎟ . (A.15) ⎝ vs2 vt2 vq2 vx2 vs2 vs2 ⎠ vs3 vt3 vq3 vx3 vs3 vs3 13/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami The Plüker relation is the same as before. In a tableau, Eq. (A.13) is expressed as 1 4 6 1 2 3 1 2 5 1 3 5 =−3 − + . (A.16) 2 3 5 4 5 6 3 4 6 2 4 6 The left-hand side is not ordered by increasing column, but the right-hand side is a combination of ordering by both increasing row and column. The Plüker relation is expressed by a two-row tableau [27], and for an m × n matrix, the formula becomes σ(i , ... , i )[a , ... , a , c , ... , c ][c , ... , c , b , ... , b ]= 0, (A.17) 1 s 1 k i i i i l m 1 t t+1 s i <···<i ,i ···<i 1 t t+1 s where (1, ... , s) = (i , ... , i ), a , b  , c  ∈ (1, ... , n), and s = m−k +l −1 > m, t = m−k > 0. 1 s j j j These tableau representations suggest algebraic structures such as character expansion. References [1] S. Ferrara, A. F. Grillo, and R. Gatto, Ann. Phys. 76, 161 (1973). [2] R. Rattazzi, V. S. Rychkov, E. Tonni, and A. Vichi, J. High Energy Phys. 12, 031 (2008) [arXiv:0807.0004 [hep-th]] [Search INSPIRE]. [3] S. El-Showk, M. F. Paulos, D. Poland, S. Rychkov, D. Simmons-Duffin, and A. Vichi, Phys. Rev. D 86, 025022 (2012) [arXiv:1203.6064 [hep-th]] [Search INSPIRE]. [4] S. El-Showk, M. F. Paulos, D. Poland, S. Rychkov, D. Simmons-Duffin, and A. Vichi, J. Stat. Phys. 157, 869 (2014) [arXiv:1403.4545 [hep-th]] [Search INSPIRE]. [5] F. Gliozzi, Phys. Rev. Lett. 111, 161602 (2013) [arXiv:1307.3111 [hep-th]] [Search INSPIRE]. [6] F. Gliozzi and A. Rago, J. High Energy Phys. 10, 042 (2014) [arXiv:1403.6003 [hep-th]] [Search INSPIRE]. [7] F. Kos, D. Poland, D. Simmons-Duffin, and A. Vichi, J. High Energy Phys. 11, 106 (2015) [arXiv:1504.07997 [hep-th]] [Search INSPIRE]. [8] F. Kos, D. Poland, D. Simmons-Duffin, and A. Vichi, J. High Energy Phys. 08, 036 (2016) [arXiv:1603.04436 [hep-th]] [Search INSPIRE]. [9] Y. Nakayama and T. Ohtsuki, Phys. Rev. D 89, 126009 (2014) [arXiv:1404.0489 [hep-th]] [Search INSPIRE]. [10] P. Basu and C. Krishnan, J. High Energy Phys. 11, 040 (2015) [arXiv:1506.06616 [hep-th]] [Search INSPIRE]. [11] H. Shimada and S. Hikami, J. Stat. Phys. 165, 1006 (2016) [arXiv:1509.04039 [cond-mat.stat-mech]] [Search INSPIRE]. [12] I. Esterlis, A. L. Fitzpatrick, and D. M. Ramirez, J. High Energy Phys. 11, 030 (2016) [arXiv:1606.07458 [hep-th]] [Search INSPIRE]. [13] P. Dey, A. Kaviraj, and A. Sinha, J. High Energy Phys. 07, 019 (2017) [arXiv:1612.05032 [hep-th]] [Search INSPIRE]. [14] S. Hikami, arXiv:1708.03072 [hep-th] [Search INSPIRE]. [15] W. Li, arXiv:1711.09075 [hep-th] [Search INSPIRE]. [16] D. Bashkirov, arXiv:1310.8255 [hep-th] [Search INSPIRE]. [17] S. Hikami, arXiv:1801.09052 [cond-mat.dis-nn] [Search INSPIRE]. [18] A. T. James, Ann. Math. Stat. 39, 1711 (1968). [19] T. Koornwinder and I. Sprinkhuizen-Kuyper, SIAM J. Math. Anal. 9, 457 (1978). [20] F. A. Dolan and H. Osborn, Nucl. Phys. B 678, 491 (2004) [arXiv:hep-th/0309180 [hep-th]] [Search INSPIRE]. [21] Y. Nakayama, Phys. Rev. Lett. 116, 141602 (2016) [arXiv:1601.06851 [hep-th]] [Search INSPIRE]. [22] S. Rychkov and Z. M. Tan, J. Phys. A: Math. Theor. 48, 29FT01 (2015).[arXiv:1505.00963 [hep-th]] [Search INSPIRE]. [23] F. Gliozzi, J. High Energy Phys. 10, 037 (2016) [arXiv:1605.04175 [hep-th]] [Search INSPIRE]. [24] F. Gliozzi, A. L. Guerrieri, A. C. Petkou, and C. Wen, Phys. Rev. Lett. 118, 061601 (2017). 14/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami [25] F. Gliozzi, A. L. Guerrieri, A. C. Petkou, and C. Wen, J. High Energy Phys. 04, 056 (2017) [arXiv:1702.03938 [hep-th]] [Search INSPIRE]. [26] J. Cardy, J. Phys. A: Math. Theor. 46, 494001 (2013) [arXiv:1302.4279 [cond-mat.stat-mech]] [Search INSPIRE]. [27] W. Bruns and V. Vetter, Determinantal Rings (Springer, Berlin, 1988), Lecture Notes in Mathematics Vol. 1327. [28] G. Parisi and N. Sourlas, Phys. Rev. Lett. 46, 871 (1981). [29] D. C. Brydges and J. Z. Imbrie, Ann. Math. 158, 1019 (2003). [30] J. A. Gracey, Phys. Rev. D 92, 025012 (2015) [arXiv:1506.03357 [hep-th]] [Search INSPIRE]. 15/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Progress of Theoretical and Experimental Physics Oxford University Press

Conformal bootstrap analysis for the Yang–Lee edge singularity

Free
15 pages

Loading next page...
 
/lp/ou_press/conformal-bootstrap-analysis-for-the-yang-lee-edge-singularity-Kt2DdD3CbM
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the Physical Society of Japan.
ISSN
2050-3911
eISSN
2050-3911
D.O.I.
10.1093/ptep/pty054
Publisher site
See Article on Publisher Site

Abstract

Prog. Theor. Exp. Phys. 2018, 053I01 (15 pages) DOI: 10.1093/ptep/pty054 Conformal bootstrap analysis for the Yang–Lee edge singularity S. Hikami Mathematical and Theoretical Physics Unit, Okinawa Institute of Science and Technology Graduate University, Okinawa, Onna 904-0495, Japan E-mail: hikami@oist.jp Received January 18, 2018; Revised April 4, 2018; Accepted April 16, 2018; Published May 30, 2018 ................................................................................................................... The Yang–Lee edge singularity is investigated by the determinant method of the conformal field theory. 3 × 3 minors are used for the evaluation of the scale dimension. The agreement with the Padé of  expansion in the region 3 < D < 6 is improved. The critical dimension D , where the scale dimension of scalar  is vanishing, is used for the improvement of the Padé. For the understanding of the intersection point of zero loci of 3 × 3 minors, 2 × 2 minors are investigated in detail; these are connected through Plüker relations. ................................................................................................................... Subject Index I10 1. Introduction The conformal field theory was developed a long time ago [1], and the modern numerical approach was initiated in Ref. [2]. Recent studies using this conformal bootstrap method have obtained some remarkable results for the 3D Ising model [3,4], Yang–Lee edge singularities [5,6], O(N ) models [7–10], and self-avoiding walks [11]. A brief summary of the determinant method for the conformal bootstrap theory is as follows. The conformal bootstrap theory is based on the conformal group O(D,2), and the conformal block G ,L is the eigenfunction of the Casimir differential operator D . The eigenvalue of this Casimir operator is C : D G = C G , 2 ,L 2 ,L C = [( − D) + L(L + D − 2)]. (1) The solutions of the Casimir equation have been studied in Refs. [18–20]. The conformal block G (u, v) has two variables u and v, which denote the cross ratios, u = (x x /x x ) , ,L 12 34 13 24 v = (x x /x x ) , where x = x − x (x is a 2D coordinate). They are expressed as u = zz ¯ and 14 23 13 24 ij i j i v = (1 − z)(1 −¯z). For the particular point z =¯z, the conformal block G (u, v) for the spin-zero ,L (L = 0) case has a simple expression: /2 2 2 z    D D + 1 D z G (u, v)| = F , , − + 1; ,  − + 1; . (2) ,0 z=¯z 3 2 1 − z 2 2 2 2 2 2 4(z − 1) The conformal bootstrap determines  by the condition of the crossing symmetry x ↔ x .For 1 3 practical calculations, the point z =¯z = 1/2 is chosen and, by the recursion relation derived from the Casimir equation of Eq. (1), G (u, v) at z =¯z = 1/2 can be obtained [3]. ,L © The Author 2018. Published by Oxford University Press on behalf of the Physical Society of Japan. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 Funded by SCOAP by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami The conformal bootstrap analysis with a small matrix size has been investigated for the Yang–Lee edge singularity with accurate results for the scale dimensions by Gliozzi and Rago [5,6]. For other models, this bootstrap method for the determinant has been considered [12,15,21,22]. In this paper, we emphasize the importance of the structure of minors along the Plüker relations, as shown in the appendix, and we apply it to the Yang–Lee model for a scale dimension of  . The intersection point of the zero loci of 3 × 3 minors was evaluated and it gave a remarkable value of the critical exponent in Ref. [5]. In these 3 × 3 minors, if the value of the space dimension D is given, the value of  =  is determined from the intersection point. However, if one goes to larger minors, for instance 4 × 4or5 × 5 minors, one need the values of additional scale dimensions of the operator product expansion (OPE). In the study of such large minors, however, the values of  deviate from the  expansion [6]. This discrepancy should be improved by the determinant method or by analysis of the  expansion, since the conformal bootstrap method should be consistent with the  expansion [24,25]. Recently, consistent results were obtained in the O(N ) vector model through the Mellin amplitude [13]. We consider this discrepancy by repeating Gliozzi’s evaluation of the 3 × 3 minors [5] and find that 3 × 3 minors give accurate values that agree with the improved  expansion (see Fig. 8). The nature of the determinant method is still not known, and the convergence to the true value seems to be slow. As we mentioned before, we have to assume the values of the scale dimensions of the OPE, such as spin 4, spin 6, etc., for the large minors. Unfortunately we do not know precisely these higher spin values at present. Therefore, we concentrate on 3 × 3 minors without any assumption of the other scale dimensions of the OPE. The four-point correlation function is given by g(u, v) φ(x )φ (x )φ (x )φ (x ) = (3) 1 2 3 4 2 2 φ φ |x | |x | 12 34 and the amplitude g(u, v) is expanded as the sum of conformal blocks: g(u, v) = 1 + p G (u, v). (4) ,L ,L ,L The crossing symmetry of x ↔ x implies 1 3 φ φ v G (u, v) − u G (v, u) ,L ,L p = 1. (5) ,L φ φ u − v ,L The minor method consists of the derivatives at the symmetric point z =¯z = 1/2 of Eq. (5). √ √ By a change of variables z = (a + b)/2, z ¯ = (a − b)/2, derivatives are taken about a and b. Since the number of equations becomes larger than the number of truncated variables , we need to consider the minors for the determination of the values of . The matrix elements of minors are expressed by φ φ v G (u, v) − u G (v, u) (m,n) ,L ,L m n f = ∂ ∂ | (6) a=1,b=0 ,L a b φ φ u − v and the minors of 2 × 2, 3 × 3, e.g., d , d , are determinants such as ij ijk (m,n) (m,n) d = det f , d = det f , (7) ij ijk ,L ,L 2/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami where i, j, k are numbers chosen differently from (1, ... , 6), following the concise correspondence to (m, n) as 1 → (2, 0),2 → (4, 0),3 → (0, 1),4 → (0, 2),5 → (2, 1), and 6 → (6, 0). In the Yang–Lee case, a fusion rule is simply written as [5] [ ]×[ ]= 1 +[ ]+[D,2]+[ ,4] + ··· . (8) φ φ φ 4 The scalar (L = 0) term in the bootstrap equation of Eq. (6) is denoted as φ φ v G (u, v) − u G (v, u) ,0 ,0 vs0 = . (9) φ φ u − v We use the notation (vsk) (k is a number, related to the degree of the differential, v is a vector, and s is a scalar) for the derivatives of the (vs0) rule in Eq. (6). For L = 2 we use (vtk), and for L = 4we use (vqk). (vxk) is for L = 6. The scalar (L = 0) case is expressed by the hypergeometric function ( =  for the Yang–Lee edge singularity) −1−/2 −/2 3 × 2 2 −/2 vs0 =−2 ys + ys + ys , (10) 2 2 where D  1 D a ys = F , , − + 1; + ,  − + 1: (11) 3 2 2 2 2 2 2 2 2 8(−2 + a) and ys is derivative of ys. We consider the derivative at a point a = 1. For instance, we have 2 × 2 minors: vs1 vs3 vs2 vs3 d = det , d = det . (12) 13 23 vt1 vt3 vt2 vt3 In this Yang–Lee edge singularity, we find that the 2 × 2 minor, d , becomes zero at D = 6, and it provides exact values of  = 2. 2. 2 × 2 minors The intersection points of zero loci of 3 × 3 minors are decomposed to 2 × 2 minors at the critical dimensions D = 6 for the free theory. In the appendix (relation 1), this decomposition is shown. Among several 2 × 2 minors, d , d , d , ... , the most fundamental minors may be d and d , 12 13 23 13 23 which are made of lower derivatives of a and b. Note that, in the Yang–Lee model, the constraint =  is taken, and the 2 × 2 minor analysis corresponds to the 3 × 3 minor analysis of another model such as the Ising model, in which  =  . From the point of the truncation error in the OPE, as discussed in Ref. [15], it is interesting to start from 2 × 2 minors. The Yang–Lee singularity is connected to supersymmetry through the dimensional reduction of a branched polymer, where we have an exact relation of  =  + 1[14,16,17]. We will discuss the dimensional reduction D + 2 → D behavior in the Yang–Lee model in 2 × 2 minors. The 2 × 2 minors consist of two parameters, dimension D and the scaling dimension  , since we have  2 =  ( =  2 ) for the Yang–Lee case, due to the φ theory. φ φ 3/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 1. The intersection of zero loci of 3 × 3 minors (d , d , d , d , d ) for dimension D = 6. The 123 135 134 234 235 axes are (x, y) = ( , Q). The intersection point of the five lines is D = 6,  = 2 with Q = 8, which is φ φ decomposed to the 2 × 2 minor d . The upper fixed point is related to the minor d . 13 23 (i) Free field theory at D = 6 The zero loci of d gives exactly the scale dimension  = 2.0 at D = 6. The minors for vq, 13 φ vx, and vs (with a scalar scale dimension  ) can be zero: vs1 vq1 vs1 vx1 vs1 vs1 det = 0, det = 0, det = 0. (13) vs3 vq3 vs3 vx3 vs3 vs3 For the value  = 2, they yield Q(=  ) = 8,  = 10, and  = 4, respectively. These φ 4 6 ,  ,  ,  are determined exactly by 2 × 2 minors of d . This is due to the free theory, i.e., φ 4 6 13 no interactions between the scale dimensions. It is known that the six dimensions have algebraic identities, which are valid for a free theory in any dimension D [23]. Other minors yield different values; for instance, d gives  = 2.158 at D = 6. These two 23 φ different values of  are understood when we evaluate 3 × 3 minors, as shown in Figs. 1 and 2. There are two fixed points of zero loci of 3 × 3 minors in the six dimensions. The 3 × 3 minors are denoted by d in Eq. (7); for instance, d means ijk 123 ⎛ ⎞ vs1 vt1 vq1 ⎜ ⎟ d = det vs2 vt2 vq2 . (14) 123 ⎝ ⎠ vs3 vt3 vq3 (ii) D = 3 The zero loci of 2 × 2 minor do not provide good results except for D = 6 dimensions. This is due to the situation where the free theory breaks down in general dimensions. We have to consider the loci of 2 × 2 minors of a finite small value due to the interactions, instead of a zero value, for the non-trivial fixed point. The small non-vanishing value of the minor might have an important meaning for the bound of  . The loci of two important 2 × 2 minors d and d are shown φ 13 23 in Fig. 2, in which the value of d is changed by small values from −0.007 to 0.004, and the zero loci of d are also shown in Fig. 3 as a guide. The correct value of  is realized in some 23 φ 4/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 2. The loci of 2 × 2 minors d with values from −0.007 to 0.004 are shown; d = 0 is also shown by a 13 23 green line. The axes are (x, y) = (D,  ). Fig. 3. A close-up of the contour map in Fig. 2 near the  = 0 loci of minor d . The green line shows the φ 13 zero loci of d as a guide. The axes are (x, y) = (D,  ). 23 φ finite-value contours of d for D < 6. Near D = 6, deviation from the zero loci of d is expected 13 13 since  deviates from 2 as the  = 6 − D expansion shows: 2 − 0.5555. φ φ In Fig. 2, the finite-value curves of d converge to a curve starting from  = 0.25 (at D = 1) 13 φ to 2.0 (at D = 4). This curve resembles the correct  contour if the space dimension is shifted from D to D + 2. (The curve is translated to the right in Fig. 2 by two space dimensions by keeping the same value of  .) This dimensional reduction will be discussed in Sect. 4. The finite-value contour map of d shows a valley shape, and one line is located at D = 2 near the exact value  =−0.4, as shown in Fig. 3. In Fig. 3, the value of minor d is 0.0034, which φ 13 gives  =−0.4. The cusp point moves when the value of minor d is changed and its trace φ 13 may give the bound of  . 5/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 4. The zero loci of d[ ,  ] for different dimensions: D = 6 (dark blue), D = 5.5 (red), D = 5 φ  13 (green), D = 4.5 (yellow), and D = 4 (blue). The axes are (x, y) = ( ,  ). In Figs. 2 and 3, contour maps of d with small non-vanishing values are shown. 3. Critical dimension for  = 0 In the Yang–Lee edge singularity, the scale dimension  of φ becomes zero at some dimension between 2 and 3, since it is −0.4 in two dimensions and approximately 0.2 in three dimensions. We call this dimension the critical dimension D , on which the scale dimension  vanishes. In the c φ determinant method, it is known that even small minors give accurate values of the scale dimensions [5,6]. We discuss here small determinants of 2 × 2, 3 × 3, and 4 × 4 matrices for this critical dimension D . Figure 2 shows Z-shaped curves for zero or very small values of the loci of d . This figure can be understood when we plot  versus  for various values of fixed dimensions. The minor of d φ 13 has so far been considered as a function of D and  with the condition of  =  due to the φ  φ Yang–Lee model. However, in general we are able to consider  and  as two free parameters for d , and, for the Yang–Lee case, we put  =  . 13  φ In Fig. 4, we plot the zero loci of d[ ,  ] for different values of D. This minor then becomes φ  13 the usual one including the Ising case. A complicated contour for the zero loci of d is obtained especially around D = 4.4, where it has three solutions for  =  . This solution corresponds to the Z shape around (D, ) = (4, 1.5) in Fig. 2. In addition, we find that the line of  =  of the Yang–Lee condition goes through very near the zero loci of d , but it does not intersect. This corresponds to Fig. 2 with the finite-value loci of d around D = 3. With some value of D, the contours in Fig. 4 (bottom left) and Fig. 5 intersect at the point ( ,  ) = (0, 0). We find that this critical dimension from Fig. 5 is D = 2.6199. (15) As seen in Fig. 3, for a small finite value of d , the contour of d approaches the line  = 0 13 13 φ from both sides, above  > 0 and  < 0. The peak of the contours approaches the critical φ φ 6/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 5. A contour plot of the zero loci of d[ ,  ] for different dimensions: D = 2.6 (blue), D = 2.61 φ  13 (yellow), D = 2.6199 (green). The axes are (x, y) = ( ,  ). dimension D . We consider the derivative of d by the dimension D, and we estimate D as the c 13 c point at which this derivative becomes zero. For  = 0.000 001, we obtain D = 2.605 74. φ c The derivative of d by the space dimension D gives the approximate solution of D in the limit 13 c → 0. The minor d is proportional to  in the limit  → 0, since vs1, vs3 ∼  , and φ 12 φ φ −1 vt1, vt3 ∼  . We consider how the minors become zero in the limit  → 0. For instance, at D = 2.61, we have, for small  , 2 2 2 vs1 = 0.188 , vs2 = 0.139 , vs3 = 0.225 φ φ φ 1.055 0.667 1.078 vt1 = , vt2 = , vt3 = . (16) φ φ φ The coefficients p , p are obtained from φ t vs0 vt0 p 1 = . (17) vs1 vt1 p 0 Since vs0 =− , vt0 = 0.7185/ in the limit  → 0, we obtain a solution of p =−4, p = φ φ φ t a log (a is a constant), which gives the value of the central charge C = 0. With this choice of p = p , the central charge C becomes C =  /p = 0 for this critical dimension D . This means t (D,2) t c that the energy–momentum tensor operator T =  can be neglected since the OPE coefficient (D,2) of this operator p becomes zero. This is the well known c catastrophe, which leads to logarithmic conformal field theory [26]. When neglecting the  term, and only considering vs0 in Eq. (10), (D,2) we obtain the critical dimension D = 2.748 632 ··· . However, this value is too large compared to the expected value and might not be correct. We need further operators to get the correct value. We hope to get the correct value of the critical dimension D by another sophisticated method by taking higher operators. 4 × 4 minors The numbers of zero loci of minors are C = 6!/4!2!= 15. We investigated 6 4 the critical dimensions D by changing D between D = 2.58 and D = 2.59. Remarkably, at 7/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 6. The intersection of zero loci of minors at D = 2.589 53,  = 0. The axes are (x, y) = (Q,  ). D = 2.589 53, all 15 lines intersect at a single point in the contour map of (Q,  ) under the condition  = 0. If we take this as the value of the critical dimension D , we find the following φ c set of results: D = 2.589 53,  = 0, Q = 4.403 97,  = 3.871 27. (18) c φ This critical dimension D obtained with 4 × 4 minors is close to D = 2.6199 from the 2 × 2 c c minor d with the intersection point of  =  = 0. In Fig. 6, all loci coincide but, if we include 13  φ more relevant operators, this value may change. For the  = 0 case, in any dimension D, d can become zero. When  = 0, by the definition φ ij φ of the scale dimension  , the two-point correlation function G(r) = 1/r becomes a constant in the long-range limit r →∞. In the Yang–Lee edge singularity, the exponent of the density σ is related to  as σ = . (19) D − From this relation, σ is vanishing for  = 0. This means that the density is constant at the transition point. Several interesting systems are known in which the density is constant but a phase transition occurs. One example is a localization problem under the random potential. 4. Dimensional reduction The zero loci of the 2 × 2 minor d shows an interesting characteristic linear behavior for D < 4, which is approximated as  = (3D − 2)/5. In Fig. 7, the shift of the zero loci of d φ 13 for D → D + 2 is shown (translation of two dimensions to the right in Fig. 7). The blue line of Fig. 7 for the zero loci of d almost coincides with the red line of the Yang–Lee edge singularity analyzed by the Padé approximation. The red line represents the result of 3 × 3 minors and it can be 8/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 7. The D → D + 2 shifted line of the zero loci of d (yellow line) is shown by the blue line and is consistent with the estimated Yang–Lee line (red) by Padé analysis. The blue and red lines are almost the same for 1 < D < 5.5. approximated as 3D − 8 = . (20) This expression satisfies the exact values of D = 1( =−1), D = 2( =−0.4), and D = 6 φ φ ( = 2). This dimensional shift seems accidental and only appears in a lower truncation, but there is a dimensional reduction that has been proven rigorously. It is known that a branched polymer in D + 2 dimensions is equivalent to a D-dimensional Yang– Lee edge singularity for the critical phenomena. This dimensional reduction has been explained by supersymmetry [28]. A rigorous mathematical proof of this dimensional reduction was shown in Ref. [29]. We will study this dimensional reduction in a separate paper using the conformal bootstrap in a determinant method [14]. The Yang–Lee edge singularity requires  (=  2 ) =  . For a branched polymer due to dimensional reduction there is a relation  =  + 1. This relation is a manifestation of the supersymmetry of the system [16]. Since the blue line is very close to the red line in Fig. 7, one can make an estimation of the critical dimension D , in which  = 0. The intersection of d with the  = 0 line can be evaluated c φ 13 φ very precisely as D = 0.599 547 1444. Adding 2 to this value for the dimensional reduction, it gives the estimation of the critical dimension D as D = 2.599 547 1444, which is very close to other c c estimations of this paper. We discussed this value in the previous section as D = 2.6199 as in Eq. (18). 5. 3 × 3 minors and Padé analysis Up to now, we have mainly discussed 2×2 minors. If we take a spin-4 operator and its scale dimension Q =  , we need to do an analysis of the intersection of the zero loci of the 3 × 3 minors d . 4 ijk 9/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Table 1. Estimated scale dimensions by zero loci of three dimensional minors. Comparison with Pade analysis is shown. The value of parenthesis is estimation from other zero loci nearby. Q  (Padé) φ φ D = 3.0 0.174 343 (0.187 825) 4.341 06 (3.771 24) 0.229 95 D = 3.5 0.499 401 (0.500 969) 5.041 95 (4.375 56) 0.531 53 D = 4.0 0.823 283 (0.815 623) 5.711 52 (4.922 83) 0.831 75 D = 4.5 1.137 55 (1.123 71) 6.333 95 (5.446 45) 1.1300 D = 5.0 1.438 07 (1.419 87) 6.917 16 (5.959 72) 1.4255 D = 5.5 1.724 69 (1.746 82) 7.469 85 (6.466 72) 1.7165 D = 6.0 2.0 8.0 2.0 From the formula of the minors in a 3 × 6 matrix in the appendix, we have a Plüker formula such as (Eq. (A.12)) [126][345]−[123][456]+[124][356]−[125][346]= 0. (21) For instance, at D = 3, we find that the zero loci of d =[123], d =[126], d =[124], and 123 126 124 d =[346] intersect at a point, and the above Plüker formula is satisfied. In Table 1, we present the intersection of three zero loci of minors d , d , and d . The 123 134 124 value of  is obtained from the intersection point of the zero loci of the minors. However, the intersection point is not the only one, and we find at least three different intersection points for each fixed dimension D. In Table 1, we present in parentheses such nearby different intersection points of the three zero loci of the minors. In the  = 6 − D expansion,  is known up to the four-loop level [30]: 2 3 4 = 2 − 0.555 55x − 0.029 4925x + 0.021 845x − 0.039 4773x , (22) where x =  = 6 − D. Including the critical dimension D , for which  is vanishing, the above c φ expansion becomes 2 2 0.55555 = (6 − x − D ) + − x + O(x ) . (23) φ c 6 − D (6 − D ) 6 − D c c c Inserting the value of D = 2.6199 from Sect. 4, it becomes = (3.3801 − x)[0.591 70 + 0.010 695x − 0.005 561 25x 3 4 + 0.004 817 55x − 0.010 2541x + ···]. (24) This expansion is approximated by the [2,2] Padé method, a + a x + a x 0 1 2 = (3.3801 − x) , (25) 1 + b x + b x 1 2 with a = 0.591 70, a = 2.3916, a = 1.0090, b = 4.023 85, b = 1.6419. The curve of this 0 1 2 1 2 Padé is incorporated in Fig. 8. For the values of  of the Yang–Lee singularity in D dimensions, the previous results obtained by Gliozzi [5] are very close to the values of Table 1; for instance, at D = 4,  = 0.823. In another evaluation with more primary operators [6], the values  = 0.8466 in D = 4 and  = 1.455 in φ φ 10/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Fig. 8.  is estimated by a [2,2] Padé with D = 2.6199. The dots are the values from Table 1 fora3 × 3 φ c minor analysis of  . D = 5 are obtained, which are larger than the values in Table 1, and a comparison with  expansion shows the apparent deviations, as shown in Fig. 5 of Ref. [6]. Our new analysis of Padé with a fixed value at the critical dimension  = 0 seems more precise, and the  expansion and bootstrap determinant method agree well with each other, as shown in Fig. 8. One of the aims of this paper was to clarify this point by the evaluation of the critical dimension D . Unfortunately, what we have done is a numerical estimation of D by the bootstrap determinant method and it may be an approximation. It is important to obtain an exact value of D , and the comparison with  expansion will be improved with an exact value of D . 6. Summary We have considered theYang–Lee edge singularity in this article by the bootstrap determinant method, initiated by Gliozzi. Although our results for the scale dimension  are same as the 3×3 determinant result of Gliozzi [5], we have improved the consistency with the result of the  expansion. We find that the basic 2 × 2 minor d is important for the determination of the above scale dimensions through the analysis of the intersection points of dimensions one to six, although we used the practical value of  obtained by 3 × 3 minors. We have shown that the value obtained by the intersection of the zero loci of 3 × 3 minors agrees with the result of the  expansion. The discrepancy between the expansion and the determinant method is reduced with the new Padé analysis with the introduction of the critical dimension D . We obtained the critical dimension D from the minor d for the  = 0 and  → 0 limits. The c 13 φ intersection of the zero loci of d[D,  ,  ] at the point  =  = 0 gives the critical dimension φ  13  φ D = 2.6199. We have estimated the critical dimension approximately by other methods, by the peak approaching the finite d (D = 2.6050), the dimensional reduction value (D = 2.5995), and 4 × 4 13 c c minor analysis (D = 2.589). We emphasize that the estimation of the critical dimension D is useful in practical terms for the precise analysis of the Yang–Lee edge singularity between two and six dimensions by the Padé analysis of the  expansion [30]. We have discussed the dimensional reduction property in the Yang–Lee model, using the fact that a branched polymer in D + 2 dimensions is equivalent to a D-dimensional Yang–Lee edge singularity. Acknowledgements The author is grateful to Ferdinando Gliozzi for discussions on the determinant method and the useful suggestion of the critical dimension. He also thanks Edouard Brézin for discussion of the dimensional reduction problem. 11/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami This work is supported by a Japan Society for the Promotion of Science (JSPS) KAKENHI Grant-in-Aid 16K05491. The Mathematica11 system of wolfram.com is acknowledged for this research. Funding Open Access funding: SCOAP . Appendix. Plüker formula Relation 1 (3 × 3 minors) For the determinant d , which is defined as ⎛ ⎞ vs1 vs2 vs3 ⎜ ⎟ d = det vt1 vt2 vt3 ⎝ ⎠ vq1 vq2 vq3 vs1 vs2 vs1 vs3 vs2 vs3 = (vq3)det − (vq2)det + (vq1)det , (A.1) vt1 vt2 vt1 vt3 vt2 vt3 if the first determinant (d ) and second minor (d ) on the right-hand side are zero, then the third 12 13 minor (d ) should be vanishing when d = 0. 23 123 Relation 2 (2 × 4 matrix) The 2 × 4 matrix is denoted as x x x x 11 12 13 14 M = . (A.2) x x x x 21 22 23 24 The Plüker relation is [12][34]−[13][24]+[14][23]= 0, (A.3) where the minor is denoted as [ij]= (x )(x ) − (x )(x ). The application of this formula to minors 1i 2j 1j 2i of our case can be taken as vs1 vs2 vs3 vs4 M = . (A.4) vt1 vt2 vt3 vt4 Thus, in our notation for 2 × 2 minors, d =[12], d =[13], d =[14], d =[23], d =[24], d =[34] (A.5) 12 13 14 23 24 34 and we have an identity d d − d d + d d = 0, (A.6) 12 34 13 24 14 23 which is represented by the tableaux 1 2 1 3 1 4 − + = 0. (A.7) 3 4 2 4 2 3 12/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami Note that the third tableau is not ordered by increasing row and column (4 is larger than 3). Another application of this formula is to take the following matrix: vs1 vt1 vq1 vx1 M = . (A.8) vs3 vt3 vq3 vx3 In the D = 6 case, we find that 2 × 2 minors made of the above matrix are vanishing, [12]=[13]= [14]= 0 for  = 2,  = 8, and  = 10. Therefore they satisfy the relation of Eq. (A.3). In this φ 4 6 D = 6 case, we also find that, for such values of  = 2,  = 8,  = 10, the other three sets of φ 4 6 minors are vanishing at D = 6: vq1 vx1 vt1 vx1 [34]= det = 0, [24]= det = 0, vq3 vx3 vt3 vx3 vt1 vq1 [23]= det = 0. (A.9) vt3 vq3 Relation 3 (3 × 6 matrix) (i) The 3 × 6 matrix in the bootstrap minor method is ⎛ ⎞ vs1 vs2 vs3 vs4 vs5 vs6 ⎜ ⎟ vt1 vt2 vt3 vt4 vt5 vt6 . (A.10) ⎝ ⎠ vq1 vq2 vq3 vq4 vq5 vq6 The Plüker relation is obtained by a mutual exchange of numbers: [146][235]+[124][356]−[134][256]+[126][345]−[136][245]+[123][456]= 0 (A.11) This identity is obtained from the first term to the second term by (2 ↔ 6), and from the second to third by (2 ↔ 3), etc. with the sign. We can also use the following relations (obtained similarly): [126][345]−[123][456]+[124][356]−[125][346]= 0 [136][245]+[123][456]+[134][256]−[135][246]= 0; (A.12) from these equations, Eq. (A.11) becomes [146][235]=−3[123][456]−[125][346]+[135][246]. (A.13) In our minor examples, we have d d =−3d d − d d + d d . (A.14) 146 235 123 456 125 346 135 246 The terms on the right-hand side are ordered by increasing row and column. (ii) Another application can be taken as ⎛ ⎞ vs1 vt1 vq1 vx1 vs1 vs1 ⎜ ⎟ . (A.15) ⎝ vs2 vt2 vq2 vx2 vs2 vs2 ⎠ vs3 vt3 vq3 vx3 vs3 vs3 13/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami The Plüker relation is the same as before. In a tableau, Eq. (A.13) is expressed as 1 4 6 1 2 3 1 2 5 1 3 5 =−3 − + . (A.16) 2 3 5 4 5 6 3 4 6 2 4 6 The left-hand side is not ordered by increasing column, but the right-hand side is a combination of ordering by both increasing row and column. The Plüker relation is expressed by a two-row tableau [27], and for an m × n matrix, the formula becomes σ(i , ... , i )[a , ... , a , c , ... , c ][c , ... , c , b , ... , b ]= 0, (A.17) 1 s 1 k i i i i l m 1 t t+1 s i <···<i ,i ···<i 1 t t+1 s where (1, ... , s) = (i , ... , i ), a , b  , c  ∈ (1, ... , n), and s = m−k +l −1 > m, t = m−k > 0. 1 s j j j These tableau representations suggest algebraic structures such as character expansion. References [1] S. Ferrara, A. F. Grillo, and R. Gatto, Ann. Phys. 76, 161 (1973). [2] R. Rattazzi, V. S. Rychkov, E. Tonni, and A. Vichi, J. High Energy Phys. 12, 031 (2008) [arXiv:0807.0004 [hep-th]] [Search INSPIRE]. [3] S. El-Showk, M. F. Paulos, D. Poland, S. Rychkov, D. Simmons-Duffin, and A. Vichi, Phys. Rev. D 86, 025022 (2012) [arXiv:1203.6064 [hep-th]] [Search INSPIRE]. [4] S. El-Showk, M. F. Paulos, D. Poland, S. Rychkov, D. Simmons-Duffin, and A. Vichi, J. Stat. Phys. 157, 869 (2014) [arXiv:1403.4545 [hep-th]] [Search INSPIRE]. [5] F. Gliozzi, Phys. Rev. Lett. 111, 161602 (2013) [arXiv:1307.3111 [hep-th]] [Search INSPIRE]. [6] F. Gliozzi and A. Rago, J. High Energy Phys. 10, 042 (2014) [arXiv:1403.6003 [hep-th]] [Search INSPIRE]. [7] F. Kos, D. Poland, D. Simmons-Duffin, and A. Vichi, J. High Energy Phys. 11, 106 (2015) [arXiv:1504.07997 [hep-th]] [Search INSPIRE]. [8] F. Kos, D. Poland, D. Simmons-Duffin, and A. Vichi, J. High Energy Phys. 08, 036 (2016) [arXiv:1603.04436 [hep-th]] [Search INSPIRE]. [9] Y. Nakayama and T. Ohtsuki, Phys. Rev. D 89, 126009 (2014) [arXiv:1404.0489 [hep-th]] [Search INSPIRE]. [10] P. Basu and C. Krishnan, J. High Energy Phys. 11, 040 (2015) [arXiv:1506.06616 [hep-th]] [Search INSPIRE]. [11] H. Shimada and S. Hikami, J. Stat. Phys. 165, 1006 (2016) [arXiv:1509.04039 [cond-mat.stat-mech]] [Search INSPIRE]. [12] I. Esterlis, A. L. Fitzpatrick, and D. M. Ramirez, J. High Energy Phys. 11, 030 (2016) [arXiv:1606.07458 [hep-th]] [Search INSPIRE]. [13] P. Dey, A. Kaviraj, and A. Sinha, J. High Energy Phys. 07, 019 (2017) [arXiv:1612.05032 [hep-th]] [Search INSPIRE]. [14] S. Hikami, arXiv:1708.03072 [hep-th] [Search INSPIRE]. [15] W. Li, arXiv:1711.09075 [hep-th] [Search INSPIRE]. [16] D. Bashkirov, arXiv:1310.8255 [hep-th] [Search INSPIRE]. [17] S. Hikami, arXiv:1801.09052 [cond-mat.dis-nn] [Search INSPIRE]. [18] A. T. James, Ann. Math. Stat. 39, 1711 (1968). [19] T. Koornwinder and I. Sprinkhuizen-Kuyper, SIAM J. Math. Anal. 9, 457 (1978). [20] F. A. Dolan and H. Osborn, Nucl. Phys. B 678, 491 (2004) [arXiv:hep-th/0309180 [hep-th]] [Search INSPIRE]. [21] Y. Nakayama, Phys. Rev. Lett. 116, 141602 (2016) [arXiv:1601.06851 [hep-th]] [Search INSPIRE]. [22] S. Rychkov and Z. M. Tan, J. Phys. A: Math. Theor. 48, 29FT01 (2015).[arXiv:1505.00963 [hep-th]] [Search INSPIRE]. [23] F. Gliozzi, J. High Energy Phys. 10, 037 (2016) [arXiv:1605.04175 [hep-th]] [Search INSPIRE]. [24] F. Gliozzi, A. L. Guerrieri, A. C. Petkou, and C. Wen, Phys. Rev. Lett. 118, 061601 (2017). 14/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018 PTEP 2018, 053I01 S. Hikami [25] F. Gliozzi, A. L. Guerrieri, A. C. Petkou, and C. Wen, J. High Energy Phys. 04, 056 (2017) [arXiv:1702.03938 [hep-th]] [Search INSPIRE]. [26] J. Cardy, J. Phys. A: Math. Theor. 46, 494001 (2013) [arXiv:1302.4279 [cond-mat.stat-mech]] [Search INSPIRE]. [27] W. Bruns and V. Vetter, Determinantal Rings (Springer, Berlin, 1988), Lecture Notes in Mathematics Vol. 1327. [28] G. Parisi and N. Sourlas, Phys. Rev. Lett. 46, 871 (1981). [29] D. C. Brydges and J. Z. Imbrie, Ann. Math. 158, 1019 (2003). [30] J. A. Gracey, Phys. Rev. D 92, 025012 (2015) [arXiv:1506.03357 [hep-th]] [Search INSPIRE]. 15/15 Downloaded from https://academic.oup.com/ptep/article-abstract/2018/5/053I01/5025800 by Ed 'DeepDyve' Gillespie user on 21 June 2018

Journal

Progress of Theoretical and Experimental PhysicsOxford University Press

Published: May 30, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off