The weak interaction is strictly lefthanded. Parity Violating Energy Difference (PVED) experiments seek a measurable energy divergence between lefthanded and righthanded molecules from weak interaction Z^{0} neutral current exchange between nucleus and electrons. Optimistic PVED is 8·10^{12} eV. Room temperature energy background kT = 0.0257 eV. Carboncarbon bond strength is 3.6 eV. Benzil PVED H_{(fusion)} will not exceed 4·10^{10} J/gram from the weak interaction.
Mendeleev Commun. 13(3) 129 (2003)
Angew. Chem. Int. Ed. 41(24) 4618 (2002)If gravitation is parityodd, like the ChernSimons term added to EinsteinHilbert action in quantized gravitations or teleparallel gravitation in Weitzenböck space, calculated H_{(fusion)} between lefthanded and righthanded benzil crystals is 8.99 J/gram or 8% energy/mass difference for a 10^{13} difference/average Equivalence Principle mass/mass anomaly. EP parity violation arises from interaction with a measurably chiral vacuum background. A two day experiment has profound consequences for physics, chemistry, and biology. Somebody should look.
EinsteinCartan theory
EinsteinCartan theory revisited
Completely asymmetric Cartan tensor
Overview Table I. Equivalence Principle Tests EP Parity Calorimetry Experiment Figure 1. Geometry of Parity Violation Table II. Eötvös vs. Parity Calorimetry Experiments Table III. Test Mass Property Magnitudes Geometric Parity Divergence Figure 2. Crystalline Benzil Helix Repeat Unit Table IV. Benzil Differential Parity Enthalpy of Fusion Net Active Mass Summary Background Table V. Postulated Gravitation Independence Table VI. Solar Gravitation at Earth Orbit Table VII. Horizontal Acceleration vs. Latitude Parity Calorimetry Experiment Details PARITY EÖTVÖS EXPERIMENT Detail
Any credible theory of gravitation  classical or quantized  must postulate, derive, or ignore the Equivalence Principle (EP):
EITHER...
The EP is true, gravitation is parityeven gerade, inertial and gravitational masses are inseparably coupled, the vacuum is achiral. Parity violations (e.g., the Weak Interaction) are extrinsic symmetry breakings.OR
The EP is not true, gravitation is parityodd ungerade, inertial and gravitational masses can be decoupled, the vacuum is chiral (e.g., possessing a background pseudoscalar field in the massed sector that differentially interacts with opposite parity atomic mass configurations). Parity violations are intrinsic (space is a left foot) and can be demonstrated with chemically and macroscopically identical, opposite parity mass distributions (left and right shoes).
General Relativity postulates the EP. String theory derives it from stateoperator correspondence and BRST invariance of graviton vertex operators. Symmetry of the Einstein curvature tensor and contingent energymomentum tensor prohibit relativistic exchange of spin and orbital angular momenta. Detecting spinorbit coupling in binary pulsar PSR J07373039A/B requires ~20 years of observation.
Conservation of angular momentum enforced by isotropic vacuum and Noethers' theorem would not obtain for opposite parity test masses. (Parity is not a Noetherian symmetry for being discontinuous and not approximated by a Taylor series.)
Contemporary Eötvös balance studies on chemical compositions (Adelberger) are sensitive to 5·10^{14} difference/average over a three month run. Cryogenic 2.2°K Eötvös apparatus sensitive to 10^{14} is being debugged (Newman). Gravitation theory is a geometry of mass distribution disregarding chemical composition. All chemical compositions empirically vacuum free fall identically.
A left foot can only be detected by a right shoe; socks and left shoes will fit almost identically. Do (metaphoric) left and right shoes fall identically? Somebody should look. A parity calorimetry test offers a 33,000fold improvement in EP anomaly sensitivity in only two days of measurements.
Year  Investigator  Accuracy  Method 

500?  Philoponus[20]  "small"  Drop Tower 
1585  Stevin[19]  5·10^{2}  Drop Tower 
1590?  Galileo[2]  2·10^{2}  Pendulum, Drop Tower 
1686  Newton[3]  10^{3}  Pendulum 
1832  Bessel[21]  2·10^{5}  Pendulum 
1910  Southerns[22]  5·10^{6}  Pendulum 
1918  Zeeman[23]  3·10^{8}  Torsion Balance 
1922  Eötvös[24]  5·10^{9}  Torsion Balance 
1923  Potter[25]  3·10^{6}  Pendulum 
1935  Renner[26]  2·10^{9}  Torsion Balance 
1964  Dicke,Roll,Krotkov[27]  3·10^{11}  Torsion Balance 
1972  Braginsky,Panov[28]  10^{12}  Torsion Balance 
1976  Shapiro, et al.[29]  10^{12}  Lunar Laser Ranging 
1981  Keiser,Faller[30]  4·10^{11}  Fluid Support 
1987  Niebauer, et al.[31]  10^{10}  Drop Tower 
1989  Heckel, et al.[32]  10^{11}  Torsion Balance 
1990  Adelberger, et al.[33]  10^{12}  Torsion Balance 
1999  Baeßler, et al.[34]  5·10^{13}  Torsion Balance 
2008  Adelberger, et al. PDF  5·10^{14}  Torsion Balance 
201?  Schwartz  2·10^{16}  Parity Torsion Balance 
201?  Schwartz  3·10^{18}  Parity Calorimetry 
201?  MiniSTEP[35]  10^{17}  Earth Orbit 
All compositions of matter validate the Equivalence Principle (EP). Empty spacetime is isotropic to massless photons locally and astronomically: vacuum refractive index = 1; no dispersion, no dichroism, no gyrotropy. Parityviolating massed sectors are neither required nor forbidden in metricaffine, EinsteinCartan, teleparallel, and Riemannian geometry gravitations. Ashtekar has a parity violating term with the Immirzi coefficient.
An EP parity violation cannot originate in Newtonian gravitation (e.g., Green's function), metric gravitation (EP), or string theory (BRST invariance). Its falsifying presence is testable. The relevant probe of spacetime geometry is test mass geometry. Theory predicts what it is told to predict. Somebody should look.
Chemically identical, opposite parity mass distributions have opposite chirality in all directions (all atom coordinates are signreversed). They have unequal insertion energies into a chiral vacuum background  a left foot fitted with left and right shoes. That divergence ends when atom positions are randomized: melt, vaporize, dissolve, or burn. Energies of transition must be different for opposite parity insertions becoming identical achiral states. A pair of calorimeters can falsify GR and string theory.
The Earth inertially accelerates about its spin axis as it gravitationally accelerates in its solar orbit. Chiral vacuum interaction, (mass_{inertial}  mass_{gravitational}) divergence, will be modulated by shifting phase angles of inertial and gravitational acceleration with local time of day compared with opposite parity test masses' geographic orientation.
Chirality vanishes at lengths smaller than a screw's pitch. The smallest possible chiral emergent scale and densest selfsimilar lattice packing are desired. alphaQuartz SiO_{3} chirality emergent scale is within a 0.304 nm diameter sphere. Benzil (C=O)(C=O) torsion angle is within a 0.313 nm diameter sphere.
Quartz' atoms are densely packed, 79.62 atom/nm^{3}. gammaGlycine's atoms are very densely packed, 127.1 atoms/nm^{3}, enantiomorphic space groups P3_{1} and P3_{1}. Both are suitable for parity Eötvös experiments. Benzil's atoms are densely packed, 93.46 atoms/nm^{3} for a parity calorimetry experiment.
Benzil (parity calorimetry experiment) and alphaquartz (both in enantiomorphic space groups P3_{1}21 and P3_{2}21) calculate as sweet spots given M. Petitjean, "On the root mean square quantitative chirality and quantitative symmetry measures", J. Math. Phys. 40, 4587 (1999). gammaglycine (parity Eötvös experiment) in enantiomorphic space groups P3_{1} and P3_{2} also passes QCM calculation with CHI asymptotic to 1, COR =1, DSI = 0.
Two local differential scanning calorimeters located between 40°50° latitude (optimal 44.95° latitude; WGS 84) preferably between 06 October and 01 April (optimal 03 January) are abutted and positioned so that their sample pans are located along a northsouth line. Each holds a ~3 mm diameter ~17 mg solid single crystal sphere of benzil, one in space group P3_{1}21 (righthanded) and one in P3_{2}21 (lefthanded). H_{(fusion)} for both are simultaneously run. The procedure is run with new crystals at 0600 hrs, noon, 1800 hrs, and midnight local time. Halfhour intervals would fill in the curve. If all H_{fusion} at all times are not equal within experimental error (differential output would be maximum signal, null, maximum, null), the experiment is repeated the next day with the calorimeters aligned eastwest to confirm. The H_{fusion} will have a six hour phase shift on the second day if the signal is real.
Metric gravitation demands the two numbers must always be identical, or H_{fusion} always equals zero. If there is a reproducible nonzero H_{fusion} General Relativity (GR) was founded upon two empirically falsified postulates  the Equivalence Principle (EP) and the isotropy of space. Affine, teleparallel, and noncommutative gravitation are validated.
There are four possible outcomes for a parity calorimetry experiment in benzil:
Differing (transition energy)/gram, is immediately testable. Allow both the DSCs to cool to obtain both H_{crystallization}. Reheat to obtain both H_{fusion} again. If the net nonzero signal originated from opposite parity mass distributions  scrambled by melting  the enthalpies of crystallization and refusion will be identical/mass in each calorimeter and between calorimeters. The initially observed divergence arose from opposite parity mass distribution geometry.
An observed net signal is further validated by running each day's calorimeter orientation at hourly or shorter intervals to fill in the sine curve response of H_{fusion} versus time of day. Those two plots would constitute inarguable evidence for a chiral vacuum background and Equivalence Principle parity violation. Helicity, optical chirality, and geometric chirality are insufficient.
Petitjean's rigorous parity divergence mathematics can be relaxed by employing benzil single crystal needles with wildly different moments of inertia rather than solid spheres. A 2_{1} screw axis is simultaneously left and righthanded. Elegant tests separately examine opposite chirality crystals of
1ferrocenyl2phenylethanedione
space group P2_{1}2_{1}2_{1}
Acta Cryst. C52 773 (1996)
and then opposite parity crystals of
1,1'ferrocenediylbis(2phenylethanedione)
space groups P3_{1}21 and P3_{2}21
Acta Cryst. C52 2465 (1996)
Their chemical compositions are essentially identical but their symmetries are wildly different.
Parameter  Composition Eötvös  Parity Calorimetry  

Duration and labor  90 days; 2160 hrs plus prep.  2 days; 2 hrs plus prep.  
Data workup  intensive and statistical  subtract two numbers  
Shielding and isolation  hard vacuum, temperature, magnetic, electromagnetic, vibration, mass gradient; torsion fiber relaxation... 
none  
Configuration  leveling, balance, sample join, light pressure, moments of inertia, oscillation damping,... 
align calorimeters, weigh test masses 

Precision  laser interferometer  0.1% precision adequate  
Active mass fraction  0.002398 maximum  0.9997 minimum  
Signal source  10^{13} mass/mass  8.99 joules/gram energy/mass  
Sensitivity, mass/mass  10^{13}  3·10^{18}  
Net output observed  needs new theory  teleparallel gravitation  
External verification  $2 million Eötvös balance  two commercial calorimeters 
A chiral nonrotating body translating through a medium will experience a consistent torque,
HK Moffat, Six lectures on general fluid dynamics and two on hydromagnetic dynamo theory, pp. 1756 in R Balian & JL Peube (eds), Fluid Dynamics (Gordon and Breach, 1977)
http://www.igf.fuw.edu.pl/KB/HKM/PDF/HKM_027_s.pdf
3.5 megabytes
pdf pp. 2527, calculation of the chiral case.
Opposite parity mass distributions will detect (absolutely and by contrast) any interactive vacuum background. This is not a parity violation experiment.
Optical chirality is a local spectral artifact. It does not detect opposite geometric parity mass distributions. Achiral crystals can powerfully rotate the plane of plane polarized light. AgGaS_{2} in nonpolar achiral tetragonal space group I42d (#122) has immense optical rotatory power: 522°/millimeter along [100] at 497.4 nm. Chiral crystals can have directions of zero optical rotation alphaQuartz 56.16° from crystallographic [0001] has zero optical rotation.
Composition chirality is irrelevant. All atoms are identical anonymous unit masses when in vacuum free fall. Chemical bonding is irrelevant. Magnetism is irrelevant, polarized spins or aligned orbital angular momenta.
Property  Fraction of Rest Mass 

rest mass  100% 
crystal lattice mass distribution parity divergence 
99.9775% (*Te) 99.9771% (*HgS) 99.9769% (*PdSbTe) 99.9730% (*AlPO_{4}) 99.9726% (*SiO_{2}) 99.9713% (*benzil) 99.9708% (*glycine) 
nuclear binding energy (low Z)  0.76% (_{2}He^{4}) 
neutron versus proton mass  0.14% 
electrostatic nuclear repulsion  0.06% 
electron mass  0.03% 
unpaired spin mass  0.005% (^{55}Mn**) 
nuclear antiparticle exchange  0.00001% 
Weak Force interactions  0.0000001% 
Gravitational binding energy  0.000000046% Earth*** 0.0000000019% moon 
All chirality eventually vanishes with decreasing volume elements. Given a screw, chirality vanishes at lengths smaller than its pitch. Best efforts are then homochiral atomic helices (crystallographic screw axes) with subnanometer emergent scale (0.323 nm diameter sphere for alphaquartz, 0.465 nm diameter sphere for benzil.)
Geometric parity divergence is chirality simultaneously calculated in all directions atom by atom. Atoms are anonymous and identical unit masses; only their relative distribution in 3space is of concern. Nuclear positions are empirically indistinguishable from atomic positions. Comparison of separate macroscopic bodies can only be achieved if every atom can be locally assigned coordinates relative to the others and globally compared between bodies. Single crystal test masses satisfy these requirements. Test masses must be solid spheres  no drilling, no hollowing  given the math of quantitative geometric parity divergence and to eliminate direction bias.
There are 11 pairs of opposite parity space groups of 65 chiral space groups in 230 3space periodic crystallographic space groups total. Howard Flack (e.g., the Flack parameter) is authoritative,
Helv. Chim. Acta 86 905 (2003) (pdf)
Explicit calculations demonstrate opposite parity space groups P3_{1}21 (righthanded screw axes) and P3_{2}21 (lefthanded screw axes) always give quantitative geometric parity divergence rapidly asymptotic to theoretical maximum  CHI>1; COR=1, DSI=0  with increasing radius. Slope is 2 by theory. The smaller the intercept the faster CHI>1.
J. Math. Phys. 40(9) 4587 (1999)
Chem. Mater. 15 464 (2003)
Benzil
Benzil, without hydrogens
alphaQuartz
Cinnabar
Cinnabar, mercury sublattice
Tellurium
Consider opposite parity P3_{(1,2)}21 crystal lattices whose formula units are achiral molecules. Dissolution, melting, sublimation, and combustion destroy the crystal lattices and their mass distribution parity divergence. The solid state inertialgravitational mass divergence must disappear during the phase transition  parity differential enthalpies of solution, fusion, sublimation, and combustion. Diastereotopic insertion energies into chiral vacuum are observable when parity divergence vanishes.
Benzil is an achiral molecule in solution, molten, or gas phase. Crystal lattice forces rotationally distort the molecule and stack it into homochiral helices giving either space group P3_{1}21 or space group P3_{2}21 for the solid,
Benzil melts between 9495 C, 112 joules/gram enthalpy of fusion (latent heat of melting). 8.99 joules/gram is an 8% divergence. Differential scanning calorimeters (NOT oscillatory!) have 0.1% precision. 10^{13} parity mass divergence detectable in three months is 10^{15} parity energy divergence detectable in 15 minutes.
Property[109]  Value  

Molecular weight 
210.2322 g/mol  
Triple point 
94.864°C  
Dynamic melting point 
Onset 94.43°C 
Meniscus 94.77°C 
Melt 95.08°C 
Thermodynamic melting point 
Onset 94.55°C 
Meniscus 94.72°C 
Melt 94.86°C 
Enthalpy of fusion mp = 94.82°C 
112.0 J/g 26.77 cal/g 
23.546 kJ/mol 5.6276 kcal/mol 

Enthalpy of fusion mp = 94.85°C 
110.6 J/g 26.44 cal/g 
23.26 kJ/mol 5.559 kcal/mol 

Enthalpy of fusion mp = 94.86°C 
112.0 J/g 26.76 cal/g 
23.54 kJ/mol 5.626 kcal/mol 

Differential enthalpy of parity divergence 
8% for 10^{13} g/g parity anomaly E = (10^{16} kg)(299,792,458 m/sec)^{2} E= 8.99 joules 
The largest active mass fraction composition divergence Adelberger et al. examine is nuclear binding energy of titanium vs. beryllium. Newman examines magnesium vs. beryllium. All other property contrasts are much smaller. Most of the loaded mass is inert in a composition experiment. Correcting for isotopic abundance and their respective nuclear binding energies,
J.H. Gundlach, New J. Phys. 7 205 (2005)
R. Newman, Class. Quantum Grav. 18 2407 (2001)
p = 938.271998 MeV n = 939.565330 MeV Be = 6.462844 MeV/baryon binding energy Mg = 8.265129 MeV/baryon binding energy Ti = 8.714634 MeV/baryon binding energy [Ti  Be]/[(30.9300n + 26p)/56.9300] = 0.002398 of total mass is active mass [Mg  Be]/[(17.3202n + 16p)/33.3202] = 0.001919 of total mass is active mass
Mass distribution parity divergence of atomic nuclei for benzil is 0.999713 of total mass. That is a factor of 417 better than Adelberger and 520 better than Newman. A 10^{13} sensitivity composition Eötvös experiment is a 23·10^{18} sensitivity calorimetry parity experiment for improved signal amplitude and active mass. A factor of 33,000 or 41,000 improvement is significant.
We are informed that Adelberger's Be/Ti comparison also tests baryon number (neutrons versus protons). Baryon number through Noether's theorem couples to an internal symmetry, SU(3) "winding number." Conserved quantities arising from internal symmetries cannot firstorder source an observable. Given the scholarly publication we run its numbers and contrast net active mass with a parity Eötvös experiment in quartz,
Phys. Rev. Lett. 100 041101 (2008) (PDF)
[(1.001077)  (0.99868)/(0.5)(1.001077 + 0.99868)][(8)(4.84]= 0.0928 g active mass
In alphaquartz with CHI=1 and active mass from atomic nuclear
positions only,
(0.999726)(1)[(8)(4.84)]= 38.709 g active mass
38.709/0.0928 = 417 times the active mass
3·10^{14} composition sensitivity, given the same magnitude of anomaly coupling, becomes 7·10^{17} geometric sensitivity. It cannot do worse than null. Somebody should look.
"It doesn't matter how beautiful your theory is, it doesn't matter how smart you are. If it doesn't agree with experiment, it's wrong," Richard Feynman.
A parity calorimetry experiment requires two calorimeters, one for each space group of crystal. The calorimeters are oriented geographic northsouth one day then eastwest the next day. For each orientation a paired opposite parity enthalpy of fusion of benzil is run at local 0600 hrs, noon, 1800 hrs, and midnight to cycle maximum signal, null, maximum, and null signals from the phase angle of Earth's inertial spin and gravitational orbit. A real net signal, H_{fusion}, will be sinusoidal during each day and phaseshifted 90° between days.
There is no parity Nordtvedt effect. Natural quartz is flawed and racemic, as are all chiral minerals. The Earth masses 5.9742·10^{24} kg. Wet biomass is 3.6·10^{14} kg. Water is not chiral. All chiral protein amino acids are Lconfiguration. All chiral natural sugars are Dconfiguration. Of the very small (dry chiral biomass)/(total Earth mass), meat and wood cancel.
A nonnull EP parity experiment does NOT contradict any existing observation.
Equivalence Principle (EP) composition tests contrast properties coupled to symmetries through Noether's theorem. Required were continuous symmetries or approximation by a finite or countably infinite number of independent infinitesimal generators (Taylor expansion) consistent with smooth Lie groups. Other dependencies, given a larger infinite number of generators (GR and the Bianchi identities), were acceptable.
Parity is the only external symmetry with no continuous or summed infinitesimal approximation. It is excluded from Taylor expansions, smooth Lie groups, and Noether's theorem.
The existence of a symmetry operator implies the existence of a conserved observable. Given G is the Hermitian generator of nontrivial unitary operator U (e.g., parity), then if U commutes with Hamiltonian H so does G [H,G]=0. If U commutes with H it is a symmetry and a conserved quantity. Any system that is initially in an eigenstate of U evolves over time to other eigenstates having the same eigenvalue.
U = c then,
Uexp(itH) = exp(itH) U [U commutes with H]
= exp(itH) c
= c exp(itH)
so exp(itH) is again an eigenstate of U, with the same eigenvalue c. Discrete symmetries also give conserved quantities in classical mechanics (e.g., bifurcation theory of dynamical systems). Parity the symmetry is coupled to geometric parity the property without Noether's theorem.
Class  Invariance  Conserved Quantity 

Proper orthochronous Lorentz symmetry 
translation in time (homogeneity) 
energy 
translation in space (homogeneity) 
linear momentum  
rotation in space (isotropy) 
angular momentum  
Discrete symmetry 
P, coordinates' inversion  spatial parity 
C, charge conjugation  charge parity  
T, time reversal  time parity  
CPT  product of parities  
Internal symmetry (independent 
U(1) gauge transformation  electric charge 
U(1) gauge transformation  lepton generation number  
U(1) gauge transformation  hypercharge  
U(1)_{Y} gauge transformation  weak hypercharge  
U(2) [U(1)xSU(2)]  electroweak force  
SU(2) gauge transformation  isospin  
SU(2)_{L} gauge transformation  weak isospin  
PxSU(2)  Gparity  
SU(3) "winding number"  baryon number  
SU(3) gauge transformation  quark color  
SU(3) (approximate)  quark flavor  
S((U2)xU(3)) [U(1)xSU(2)xSU(3)] 
Standard Model 
EP tests exploit external symmetries' observables. Internal symmetries' observables (gauged using fiber bundle theory, e.g., charge conjugation) transform fields amongst themselves leaving physical states (translation, rotation) invariant. A local gauge transformation always exists to make the local gaugefield vanish. Two vector potentials differing only by a gauge transformation give the same field. EP tests opposing properties coupled to internal symmetries are empirical first order default nulls.
Year  Date  Geocenter to sun, 10^{6} km  Acceleration, cm/sec^{2} 

2008  02 January  147.096603  0.613347 
2009  04 January  147.095552  0.613356 
2010  03 January  147.098040  0.613335 
2011  03 January  147.105761  0.613271 
2012  05 January  147.097207  0.613342 
2013  02 January  147.098161  0.613334 
average  147.09855  0.613331  
1 AU  03 April  149.598  0.593007 
2008  04 July  152.104160  0.573627 
2009  04 July  152.091131  0.573725 
2010  06 July  152.096448  0.573685 
2011  04 July  152.102197  0.573642 
2012  05 July  152.092424  0.573715 
2013  05 July  152.097426  0.573678 
average  152.097298  0.573679  
1 AU  03 October  149.598  0.593007 
and surface inertial centripetal acceleration (sidereal day, WGS 84;
sea level unless noted) of Earth's rotation.
r = (6378136.46)[1([sin^{2}(lat)]/298.257223563)] meters
a = (3.380199)(cosine[lat])/[1(0.006694380)cosine^{2}(lat)] cm/sec^{2}
horizontal component of a = a[sin(lat)]
Resultant, cm/sec^{2}  Geocentric Latitude, degrees  Horizontal Component, sin(lat) 
Centripetal Acceleration, cm/sec^{2} 


3.66  45  550 mph east ground speed 
commercial airliner 

1.46490  60  0.866025  1.691516  
1.58993  55  0.819152  1.940941  
1.66673  50  0.766044  2.175761  
1.69294  45  0.70711  2.394172  
1.69294  44.951894  0.706513  2.396188  
1.66770  40  0.642788  2.594484  
1.59175  35  0.573576  2.775137  
1.46736  30  0.500000  2.934715  
1.29826  25  0.422618  3.071958  
1.08960  20  0.342020  3.185778  
0.84770  15  0.258819  3.275266  
0.57993  10  0.173648  3.339705  
0.29446  5  0.087156  3.378578  
0  0  0  3.391570 
g = (978.032677)[1+(0.00193185139)sin^{2}(lat)]/[1(0.00669437999)sin^{2}(lat)] cm/sec^{2}
dg/dh = 0.000308766[1(0.0014665)sin^{2}(lat)] cm/sec^{2}meter
dg/ds = 0.0008109[cos[(2)(lat)]+(0.0022)cos[(4)(lat)](0.0033)cos[2(lat)]sin^{2}(lat)] cm/sec^{2}kmr = geocentric radius
a = total centripetal acceleration
lat = latitude
g = gravitational acceleration
h = altitude above sea level
s = distance
A 10^{13} difference/average mass/mass divergence is a (c^{2})(10^{13}) energy/mass divergence, or 8.99 joules/gram. Racemic benzil powder enthalpy of fusion is 112 joules/gram. 8.99/112 = 8%. Differential scanning calorimeters are accurate to 1% and precise to 0.1%. Benzil is an achiral molecule. Melting its crystal lattice destroys  not racemizes  its parity divergence.
Benzil must be crystallized from solution. Czochralski melt pull and BridgmanStockbarger directional solidification both give plastic deformation. Atom positions are distorted in crystal lattices.
Differential scanning calorimeters require 1015 milligram samples. That is a benzil solid sphere 2.53 mm in diameter, d=1.23g/cm^{3}. Subdued light: benzil is homolytically cleaved between its carbonyls at wavelengths of 430 nm and shorter.
On a good day, sort space groups by hemihedral facets. Otherwise, single edge razor blade to cut needles into cubes. Needle length is crystallographic caxis. View along the caxis between crossed linear polarizers to sort into lefthanded and righthand crystals. Benzil optical rotation of ±23º/mm at the sodium Dline, 589 nm.
Run trimmed cubes through a crystallographer's air grinder to shape spheres. Change the abrasive paper between lefthanded and righthand runs. On filter paper within a Buchner funnel quick ethanolwater wash and suck dry to polish the surfaces. Eight balls of each parity do the two days' runs.
Benzil has vapor pressure of 1 torr @ 128 C. To prevent heatflow errors from sublimation, encapsulate the tared sphere in conformal aluminum foil.
If inertial and gravitational masses periodically uncouple for opposite parity mass distributions their vacuum free fall minimum action trajectories will be nonparallel. One parity will fall like everything else  a left shoe or sock on a left foot. The other parity will fall differently  a right shoe on a left foot. The horizontal component will exert a periodic differential torque if they are mounted on opposite sides of a vertical rotor.
http://www.npl.washington.edu/eotwash/experiments/equivalencePrinciple/newWashPendulum.jpg
Eötvös vertical torsion balance rotor loaded with
two opposed sets of test masses and hanging from a tungsten fiber (top).
The preferred test masses are seedless solid single crystal spheres of space group P3_{1}21 and P3_{2}21 alphaquartz. The two control experiments are each quartz parity against amorphous fused silica test masses. Fused silica has a lower density than crystalline alphaquartz. Gold plating on fused silica must be marginally thicker (+a) on a marginally smaller radius (Ra) solid sphere to balance mass and moments of inertia. "a" is directly proportional to radius. For R=1, a=0.0087810
a^{3}  3Ra^{2} + 3R^{2}a  [(d_{0}  d_{1})/(d_{2}  d_{1})] = 0
a = fused silica radial decrement
R = quartz radius
d_{0} = density of quartz, 2.649 g/cm^{3} nominal
d_{1} = density of fused silica, 2.203 g/cm^{3} nominal
d_{2} = density of gold, 19.283 g/cm^{3} nominal (xray density)
Densities should be temperatureadjusted.