Arxiv:1809.06478V2 [Nucl-Th] 13 Dec 2018 Zp 2 P 2 N 2 S 2 4GE (Q ) = Qpge(Q ) + Qnge(Q ) − GE(Q )

Weak radius of the proton

C. J. Horowitz1, ∗ 1Center for Exploration of Energy and Matter and Department of Physics, Indiana University, Bloomington, IN 47405, USA (Dated: December 14, 2018) The weak charge of the proton determines its coupling to the Z0 boson. The distribution of weak charge is found to be dramatically diﬀerent from the distribution of electric charge. The proton’s weak radius RW = 1.545 ± 0.017 fm is 80% larger than its charge radius Rch ≈ 0.84 fm because of a very large pion cloud contribution. This large weak radius can be measured with parity violating electron scattering and may provide insight into the structure of the proton, various radiative corrections, and possible strange quark contributions.

The distributions of charge and magnetization provide one has, crucial insight into the structure of the proton. Hofs- tadter in the 1950s determined that the charge distribu- 2 2 Qn 2 1 2 RW = Rch + Rn − Rs . (4) tion in the proton has a signiﬁcant size ≈ 0.8 fm [1], while Qp Qp

experiments at Jefferson Laboratory have shown that the 2 Here the neutron charge radius squared is Rn = magnetization is distributed differently from the charge 2 2 [2–4]. More recently, an experiment with muonic hydro- −0.1148 ± 0.0035 fm [11]. Note that this Rn value is gen has found a surprising value for the charge radius of a somewhat old result that could be improved with a modern experiment. The nucleon’s strangeness radius the proton Rch [5] that is smaller than previous determi- 2 nations with electron scattering or conventional hydro- squared Rs is defined, gen spectroscopy [6]. This proton radius puzzle has mo- 2 d s tivated considerable theoretical and experimental work R = −6 G 2 . (5) s dq2 E q =0 and is presently unresolved [7]. In addition to its electric charge, the proton has a Zp s There have been several measurements of GE and GE weak charge Qp that characterizes the strength of the using parity violating electron scattering, see for example 0 vector coupling to the Z boson. The Jefferson Labora- 2 [12, 13]. Recently the Qweak collaboration determined Rs tory Qweak collaboration has recently measured Qp using in addition to Qp [8] (from their ρs value), parity violating electron scattering [8], R2 = −0.013 ± 0.007 fm2. (6) Qp = 0.0719 ± 0.0045 . (1) s This value provides a sensitive test of the standard model Alternatively, Lattice QCD calculations [14] find a con- at low energies, and the future P2 experiment aims to sistent but more precise value,

improve the accuracy further [9]. 2 2 In the standard model, the weak neutral current is a Rs = −0.0054 ± 0.0016 fm (LQCD), (7) mixture of the isovector weak and electromagnetic cur- where we have combined their different errors in quadra- rents. As a result the Sachs form factor that describes ture. If the Q collaboration constrain R2 to the the vector interaction of the Z0 boson with the proton, weak s Zp LQCD value, they find a slightly different value for Qp GE is related to the conventional electric form factors of instead of Eq. 1, p n the proton GE and neutron GE plus a possible strange s quark contribution GE [10], Qp = 0.0685 ± 0.0038 . (8)

arXiv:1809.06478v2 [nucl-th] 13 Dec 2018 Zp 2 p 2 n 2 s 2 4GE (q ) = QpGE(q ) + QnGE(q ) − GE(q ). (2) We note that Eq. 8 is consistent with Eq. 1 within errors Here q2 = −q2 > 0 is the square of the momentum trans- and both are consistent with the standard model value, µ see below. fer and Qn is the weak charge of the neutron. The distribution of weak charge in the proton is charac- We list in Table I values of the proton weak radius terized by a size or root mean square radius R that fol- from Eq. 4. These involve the large ratio of the neu- W tron to proton weak charges. At tree level Q = −1 and lows from the q2 dependence of the form factor 4GZp(q2). n E Q = 1 − 4 sin2 Θ ≈ 0.05. Including radiative correc- Using p W tions yields the standard model values Qn = −0.9902 and d 1 2 Zp Qp = 0.0710 [15] . Our procedure is to use the LQCD QpRW = −6 (4GE ) , (3) dq2 q2=0

1 We neglect radiative corrections for the strange quark contribu- ∗Electronic address: [email protected] tion. 2

2 0 Zp value for Rs and the standard model values for Qp and netic coupling of the Z to the proton GM has the same Qn. This yields the most precise RW where the error is form as Eq. 2, 2 dominated by the error in Rn. If we use the Qweak value Zp 2 p 2 n 2 s 2 for Qp from Eq. 8 instead, we have a similar value for 4GM (q ) = QpGM (q ) + QnGM (q ) − GM (q ) . (9) RW with a slightly larger ≈ ±0.035 fm error. Table I has two lines for RW because of the proton radius puzzle. Since |Qn| Qp, we expect the weak magnetic radius The second line (and the line for the neutron, see below) of the proton to be close to the (conventional) magnetic corresponds to the smaller Rch from muonic hydrogen. radius of the neutron, and very diﬀerent from RW . To gain additional insight into RW Rch, we con-

Particle Rch (fm) RW (fm) ∆R (fm) sider coordinate space representations related to the lo- p 0.877 ± 0.007 [6] 1.564 ± 0.017 0.687 ± 0.017 cal charge or weak charge density. We define the func- 0.8418 ± 0.0007 [5] 1.545 ± 0.017 0.703 ± 0.017 tion ρW (r) as the three dimensional Fourier transform of n 0.8434 ± 0.0012 Zp 2 GE (q ), 208Pb 5.503 5.826 ± 0.181[22] 0.323 ± 0.181 Z d3q TABLE I: The charge radius R , weak radius R , and weak −iq·r Zp 2 ch W ρW (r) = 3 e 4GE (q ) . (10) 208 (2π) skin ∆R = RW − Rch for the proton, neutron, and Pb nucleus. In the nonrelativistic limit ρW (r) corresponds to the weak charge density of the proton. However relativis- As shown in Table I, the weak radius of the proton tic effects can complicate this interpretation. We will R is ≈ 80% larger than the charge radius R . We W ch consider a relativistic transverse density below [16]. For define the weak skin ∆R as the difference between the now ρW (r) is defined by Eq. 10. We assume simple weak and charge radii ∆R = R − R . The large value p W ch form factors G (q2) = (1 + q2/Λ2)−2 and Gn (q2) = E √ E ∆R ≈ 0.7 fm shows that weak charges are more likely 2 2 2 −2 aq (1 + q /Λ ) with the constants Λ = 12/Rch and to be found at large distances from the origin than are 2 electric charges. a = −Rn/6 chosen to reproduce mean square radii. Of course more detailed form factors can be used, see for ex- The dramatic difference R R is a major result W ch ample [17–19], but we do not expect them to make qual- of this paper. Why does the distribution of weak charge itative differences. We neglect small strange quark con- have a spatial extent that is much greater than the extent tributions. This gives a simple analytic form for ρ (r), of the (electric) charge? We first present an explanation W in terms of hadronic coordinates and then we present 2 3 n Rn 2 ΛoΛ −Λr an alternative description in terms of quark coordinates. ρW (r) = Qp + Qn Λ − 2 e . (11) Consider a virtual transition p → n + π+. The weak 6 r 8π charge of the pion Qπ+ = Qp − Qn = 1.061 is much In Fig. 1 we plot the normalized function ρW (r)/Qp. We larger than Qp. Therefore the pion “tail”, present in the see that ρW (r) has a node near r = 0.4 fm. proton at large radius, “wags the dog” and makes a very For reference, Fig. 1 also shows the function ρ(r) = large contribution to RW . Λ3e−Λr/(8π) determined from the Fourier transform of The weak radius of the proton is related to the charge p 2 the proton electric form factor GE(q ). There are large radius (squared) of the neutron. Consider a virtual tran- differences between ρW (r) and ρ(r). Note that the func- sition n → p + π−. The π− produces a negative charge tions in Fig. 1 have been multiplied by r2 to emphasize 2 distribution at large distances resulting in a negative Rn. their large r behaviors. 2 Thus Rn is negative because of the neutron’s pion cloud. We now consider a relativistically consistent represen- Another equivalent way to understand RW Rch is tation of the weak charge density of the proton provided t to consider the distribution of up and down quarks. The by the transverse density ρW (b) [16]. This is the two weak charge of a proton Qp is small because of a sensitive Zp dimensional Fourier transform of the weak F1 form fac- cancelation between the weak charges of two up quarks tor, and the weak charge of one down quark. Therefore ρW (r) is very sensitive to small differences between the distri- Z d2q ρt (b) = e−iq·b4F Zp(q2) , (12) butions of up and down quarks. If the up quarks have W (2π)2 1 a somewhat larger radius than the down quarks this will lead to RW Rch. Z Zp Zp We note that RW is sensitive to radiative corrections. t qdq 4GE + τ4GM ρ (b) = J0(qb) . (13) If one evaluates RW in Eq. 4 with the tree level weak W 2π 1 + τ charges Qn = −1 and Qp = 0.05 the result is RW ≈ 1.8 2 2 fm. This is about 15% larger than the values in Table I. Here τ = q /4MN and J0 is a cylindrical Bessel function. One should study further the impact of radiative correc- We neglect strange quark contributions and use simple n p tions on RW . parameterizations of the form factors GE = µnτGE/(1+ n p p p For completeness, we also discuss the weak magnetic 5.6τ) [20], GM = µnGE, and GM = (1 + µp)GE with p 2 2 −2 radius of the proton. The form factor describing the mag- µp = 1.7928, µn = −1.9130 and GE = (1 + q /Λ ) . 3

0.2 0.5

0.4 )

) 0.1 -1 -1 0.3 0 (r) (fm (b) (fm i i t ρ ρ

2 0.2 r -0.1 b 0.1 -0.2 0

0 0.5 1 1.5 2 0 0.5 1 1.5 2 2.5 r (fm) b (fm)

FIG. 1: (Color online) Fourier transforms of the proton elec- FIG. 2: (Color online) Impact parameter b times the trans- tric form factor r2ρ(r) (solid black line) and the normalized verse charge density of the proton bρt(b) (solid black line), 2 t weak form factor r ρW (r)/Qp (dashed blue line) versus radius transverse weak density of the proton bρW (b)/Qp (dashed r, see Eq. 10. blue line) and the transverse charge density of the neutron t bρn(b) (dotted red line), see Eq. 12.

0 Zn The impact of more accurate fits, see for example [21], coupling of the Z to the neutron is GE , should be explored. Zn 2 p 2 n 2 s 2 t 4GE (q ) = QnGE(q ) + QpGE(q ) − GE(q ). (14) Figure 2 shows the transverse weak density ρW (b)/Qp, and the transverse charge densities of the proton ρt(b), n 2 Defining the neutron weak radius from QnRW = and neutron ρt (b). These are defined as in Eq. 12 us- Zn 2 n −6d(4G )/dq | 2 yields, p n t E q =0 ing F1 and F1 . At large distances both ρW (b)/Qp and ρW (r)/Qp are large and positive. This shows the large n 2 2 Qp 2 1 2 t RW = Rch + Rn − Rs . (15) contribution of the pion tail. Furthermore, ρW (b)/Qp Q Q t n n is very different from ρ (b). Likewise ρW (r)/Qp is very t different from ρ(r). However ρW (b) is positive for small Given |Qn| Qp, the weak radius of the neutron, see impact parameters while ρW (r) is negative at small r. Table I, is very close to the charge radius of the proton n Note that small impact parameter b does not necessarily RW ≈ Rch. For light N = Z nuclei such as the deuteron, correspond to small radius r. The large differences seen 4He or 12C the weak radius of a nucleus is expected to in Figs. 1, 2 emphasize that the weak charge in a proton be close to its charge radius. For mirror nuclei such as is distributed very differently from the electric charge. 3He and 3H the weak radius of 3He should be close to the charge radius of 3H and vise versa. We summarize relativistic effects. In general, all of the Can one gain insight by comparing the weak skin of equations in this paper are fully relativistic and there the proton to the weak skin of heavy nuclei? We note are no relativistic corrections to any of our numerical re- that Hofstadter explored the charge densities of the pro- sults. The only complication from relativity is in the ton and heavy nuclei with very similar experiments. For physical interpretation of ρ (r) that is defined by the W heavy nuclei with N > Z we expect a neutron skin with three dimensional Fourier transform in Eq. 10. In order some of the extra neutrons collecting in the surface re- to probe the system at small distances, one must scat- gion so that the neutron radius is greater than the pro- ter with a momentum transfer q that is comparable to ton radius. As a result there will be a weak skin with the nucleon’s mass. This causes the nucleon to recoil R > R . This has now been verified for 208Pb, where and this recoil must be taken into account. Therefore W ch R has been measured in the PREX experiment [22, 23], ρ (r) shown in Fig. 1 can not be interpreted as a local W W see Table I. weak charge density in coordinate space. In contrast the Clearly the weak skin of the proton is not produced by transverse density ρt (b) defined in Eq. 12 and shown in W a neutron skin even though both the proton and 208Pb Fig. 2 can be interpreted as a local weak charge density have R > R . Instead, the weak skin of the proton because it is not impacted by boosting the system along W ch can be thought of as coming from an “up quark skin” the direction of q from the nucleon’s recoil. rather than a neutron skin. The up quark skin describes For comparison, we now discuss weak radii for the neu- an excess of up quarks at large radii in the proton. For tron and heavier nuclei. The form factor describing the example, a virtual π+ cloud at large radii will increase 4 the density of up quarks and reduce the net density of suggesting that very neutron rich Ca isotopes, perhaps down quarks (minus down antiquarks) and produce an up to 70Ca, are particle stable. The nucleus 70Ca, with up quark skin. Thus, the proton could also be thought 2.5 times more neutrons than protons, could have a very of as having a “pion skin”. thick weak skin. We end by exploring if there are nuclei with RW Rch In conclusion, we have calculated the distribution of that might have weak skins at least somewhat compara- weak charge in the proton and find it to be dramatically ble to the very large ∆R ≈ 0.7 fm of the proton. We different from the distribution of electric charge. The consider two possibilities. The first is a neutron halo nu- weak radius RW is ≈ 80% larger than the charge radius 11 cleus such as Li where the neutron radius and RW may Rch because of a very large pion cloud contribution. This be dominated by the weakly bound neutron halo that large weak radius probes proton structure including dif- extends to very large radii [24]. The second possibility ferences between up and down quark distributions, and is simply a very neutron rich nucleus. This nucleus will RW can be measured with parity violating electron scat- likely have a thick neutron skin (and hence weak skin) tering. because of all of the “extra” neutrons. It is interesting to consider Ca isotopes. The heaviest 40 stable N = Z nucleus is Ca where we expect Rch to be Acknowledgments slightly larger than RW (∆R < 0) because the protons are pushed out by the Coulomb interaction. The CREX experiment should accurately measure RW for the dou- We thank Bill Donnelly, Farrukh Fattoyev, Misha Gor- 48 bly closed shell neutron rich isotope Ca [25]. Here RW shteyn, Zidu Lin, Jerry Miller, Seamus Riordan, and depends on poorly constrained three neutron forces. Mi- Mike Snow for helpful discussions. This work was started croscopic chiral effective field theory calculations predict at the Mainz Institute for Theoretical Physics and we a thin weak skin for 48Ca, ∆R ≈ 0.13 fm [26] while dis- thank them for their hospitality. This material is based persive optical model calculations predict a thick skin upon work supported by the U.S. Department of En- ∆R ≈ 0.25 fm [27]. Even more neutron rich Ca isotopes ergy Office of Science, Office of Nuclear Physics under are expected to have thicker weak skins. Recently the Awards DE-FG02-87ER40365 (Indiana University) and isotopes 59Ca and 60Ca were observed at RIKEN [28] de-sc0018083 (NUCLEI SciDAC-4 Collaboration).

[1] R. Hofstadter, F. Bumiller, and M. R. Yearian, Rev. 92, 031501(R) (2015). Mod. Phys. 30, 482 (1958). [15] J. Beringer et al., Phys. Rev. D 86, 010001 (2012), [2] M. K. Jones et al. (The Jefferson Lab Hall A Collabora- [PDG]. tion), Phys. Rev. Lett. 84, 1398 (2000). [16] Gerald A. Miller, Annu. Rev. Nucl. Part. Sci. 60, 1 [3] O. Gayou et al. (Jefferson Lab Hall A Collaboration), (2010). Phys. Rev. Lett. 88, 092301 (2002). [17] Z. Ye, J. Arrington, R. J. Hill and G. Lee, Phys. Lett. B [4] A. J. R. Puckett et al., Phys. Rev. Lett. 104, 242301 777, 8 (2018). (2010). [18] S. Venkat, J. Arrington, G. A. Miller and X. Zhan, Phys. [5] R. Pohl et al., Nature 466, 213 (2010). Rev. C 83, 015203 (2011). [6] P. J. Mohr, B. N. Taylor and D. B. Newell, Rev. Mod. [19] J. J. Kelly, Phys. Rev. C 70, 068202 (2004). Phys. 80, 633 (2008). [20] S. Glaster et al., Nucl. Phys. B32, 221 (1971). [7] Randolf Pohl, Ronald Gilman, Gerald A. Miller, [21] W. M. Alberico, S. M. Bilenky, C. Giunti, and K. M. Krzysztof Pachucki, Annual Review of Nuclear and Par- Graczy, Phys. Rev. C 79, 065204 (2009). ticle Science, 63,175 (2013). [22] C. J. Horowitz et al., Phys. Rev. C 85, 032501(R) (2012). [8] D. Androic et al. (The Jefferson Lab Qweak Collabora- [23] S. Abrahamyan et al. (PREX Collaboration), Phys. Rev. tion), Nature 557, 207 (2018). Lett. 108, 112502 (2012). [9] Dominik Becker et al., arXiv:1802.04759, submitted to [24] I. Tanihata et al., Phys. Rev. Lett. 55, 2676 (1985). EPJ A (2018). [25] The CREX proposal, unpublished, available at [10] J. Erler, C. J. Horowitz, S. Mantry, and P. A. Souder, hallaweb.jlab.org/parity/prex. Ann. Rev. Nucl. Part. Sci. 64, 269 (2014). [26] G. Hagen, A. Ekstr’m, C. Forss’n, G. R. Jansen, W. [11] S. Kopecky et al., Phys. Rev. C 56, 2229 (1997). Nazarewicz, T. Papenbrock, K. A. Wendt, S. Bacca, N. [12] R. González-Jiménez,J. A. Caballero, and T. W. Don- Barnea, B. Carlsson, C. Drischler, K. Hebeler, M. Hjorth- nelly, Phys. Rev. D 90, 033002 (2014). Jensen, M. Miorelli, G. Orlandini, A. Schwenk and J. Si- [13] A. Acha et al. (HAPPEX collaboration), Phys. Rev. Lett. monis, Nature Physics 12,186 (2016). 98, 032301 (2007). [27] M. H. Mahzoon, M. C. Atkinson, R. J. Charity, and W. [14] Jeremy Green, Stefan Meinel, Michael Engelhardt, Ste- H. Dickhoff, Phys. Rev. Lett. 119, 222503 (2017). fan Krieg, Jesse Laeuchli, John Negele, Kostas Orginos, [28] O. B. Tarasov et al., Phys. Rev. Lett. 121, 022501 (2018). Andrew Pochinsky, and Sergey Syritsyn, Phys. Rev. D