Measurement of the Do Lifetime from the Upgraded Mark II Detector at ...

5 downloads 0 Views 930KB Size Report
decays Do + K-x+, K-w+x', and K-7rr+7r-rr+, we measure the Do lifetime to be 0.44?:*:: f 0.06 ..... Because of this, we took the range over which we varied rB.
SLAC - PUB LBL - 23321 May 1987 E

Measurement the Upgraded

of the Do Lifetime Mark

II Detector

- 4304

from at PEP*

S. R. Wagner,(l) D. A. Hinshaw,(l) R. A. Ong,t2) G. Abrams,c3) C. E. Adolphsen,c4) C. Akerlof,t5) J. P. Alexander,t2) M. Alvarez,(l)f(“) A. R. Baden,(3)y(b) J. B a 11am,t2) B. C. Barish,(‘) T. Barklow,c2) B. A. Barnett,c7) J. Bartelt,t2) D. Blockus,c8) G. Bonvicini,c5) A. Boyarski,c2) J. Boyer,c3) B. Brabson,(*) A. Breakstone,tg) J. M. Brom,c8) F. Bulos,t2) P. R. Burchat,t4) D. L. Burke,c2) F. Butler,t3) F. Calvino,(l)s(a) R. J. Cence,cg) J. Chapman,c5) D. Cords,c2) D. P. Coupal,c2) H. C. DeStaebler,c2) D. E. Dorfan,t4) J. M. Dorfan,c2) P. S. Drell,t3) G. J. Feldman,c2) E. Fernandez,(l)s(a) R. C. Field,c2) W. T. Ford,(l) C. Fordham,c2) R. Frey,c5) D. Fujino,c2) K. K. Gan,12) G. Gidal,t3) T. Glanzman,c2) G. Goldhaber,c3) A. Green,(2)*(G) P. G rosse-Wiesmann,(2) J. Haggerty,(3)l(d) G. Hanson,c2) R. Harr,t3) F. A. Harris,tg) C. M. Hawkes,c6) K. Hayes,c2) D. Herrup,c3) C. A. Heusch,c4) T. Himel,c2) M. Hoenk,(6) D. Hutchinson,(2) J. Hylen,17) W. R. Innes,t2) M. Jaffre,c3) J. A. Jaros,(2) I. Juricic,c3) J. A. Kadyk,c3) D. Karlen,c2) J. Kent,t4) S. R. Klein,c2) A. Koide,(g)j(e) W. Koska,(5) W. Kozanecki,c2) A. J. Lankford,t2) R. R. Larsen,c2) M. E. Levi,c2) Z. Li,c6) A. M. Litke,c4) V. Liith,c2) J. A. J. Matthews,t7) D. I. Meyer,t5) B. D. Milliken,(‘) K. C. Moffeit,c2) L. Miiller,(2)2(f) J. Nash,t2) M. E. Nelson,(6) D. Nitz,c5) H. Ogren,t8) K. F. O’Shaughnessy,(2) S. I. Parker,cg) C. Peck,c6) M. L. Perl,c2) A. Petersen,t2) M. Petradza,c5) F. C. Porter,(‘) P. Rankin,c2) K. Riles,c2) D. R. Rust,c8) H. F. W. Sadrozinski,(4) T. L. Schalk,c4) W. B. Schmidke,c3) A. S. Schwarz,(4) A. Seiden,t4) J. G. Smith,(l) A. Snyder,c8) E. Soderstrom,(6) D. P. Stoker,(7) R. Stroynowski,(‘) R. Thun,c5) G. H. Trilling,(3) R. Tschirhart,c5) R. Van Kooten,(2) H. Veltman,(5),(c) P. Voruganti,c2) P. Weber,(‘) A. J. Weinstein,t4) A. J. Weir,c6) S. Weisz,(4)j(g) S. L. White,(‘) E. Wicklund,(6) D. R. Wood,(3) and D. Y. Wuc6)

* This work was supported in part by Department of Energy contracts DEACOJ81ER40050 (CIT), DEAA03-76SFOOOlO (UCSC), DEAC02-86ER40253 (Colorado), DEAC03-83ER40103 (Hawaii), DEAC02-84ER40125 (Indiana), DEAC03-76SF00098 (LBL), DEAC02-84ER40125 (Michigan), and DEAC03-76SF00515 (SLAC), and by the National Science Foundation (Johns Hopkins).

(l) University of Colorado, Boulder, Colorado 80309 (2)Stanford Linear Accelerator Center, Stanford Uniuersity, Stanford, California 94.905 t3) Lawrence Berkeley Laboratory and Department of Physics, University of California, Berkeley, California 94 720 c4)University of California, Santa Cruz, California 95064 c5) University of Michigan, Ann Arbor, Michigan 48109 c6)California Institute of Technology, Pasadena, California 91125 c7)Johns Hopkins University, Baltimore, Maryland 21218 t8) Indiana University, Bloomington, Indiana 47405 cg)University of Hawaii, Honolulu, Hawaii 96822

ABSTRACT We present a new measurement of the Do meson lifetime based on 31 pb-’ of e+e- interaction

data taken at fi

= 29 GeV. The data were recorded with

the upgraded Mark II detector, which allowed the identification

of a clean Do

signal from this small data set. Using a sample of 53 Do candidates based on the decays Do + K-x+,

K-w+x’,

and K-7rr+7r-rr+,

we measure the Do lifetime to

be 0.44?:*:: f 0.06 picoseconds. Submitted

(a) Present address: Universidad

Autonoma

(b) Present address: Harvard University, (c) Present address: University (d) Present address: Brookhaven

to Physical Review D

de Barcelona, Bellaterra,

Cambridge,

of California, National

Spain

MA 02138

Berkeley, CA 94720

Laboratory,

Upton, NY 11973

(e) Present address: Vista Research Inc., 3600 Bayshore Road, Palo Alto, CA 94303 (f) Present address: Lab. fiir Hochenergie Physik Bern, CH-3012 Bern, Switzerland (g) Present address: CERN, CH-1211, Genkve 23, Switzerland 2

In this paper we report a new measurement of the Do meson lifetime, formed with the upgraded Mark II detector[” 31 pb-’ of e+e- interaction

per-

at PEP. This analysis is based on

data taken at ,/Z=

29 GeV. The Do mesons were

tagged by the spectator pion from the rr+Do decay of the D*+.

The same de-

cay length method used to measure the Do lifetime in a previous publication the Mark II collaboration”’

is employed here, but the components

of

of the de-

tector crucial to this analysis have changed completely, and this measurement is statistically

independent

of the previous one.

The original Mark II detector”’

was upgraded for its planned use at the SLAC

Linear Collider (SLC), w h ere the average particle momentum the hadronic jet environment

will be higher and

denser than at PEP. The PEP run, from which the

data presented in this paper is taken, was used to check out the upgraded detector and to take advantage of the many improvements

to the detector in the study of

physics at PEP energies. The new components used in this analysis are a 72 layer cylindrical

main drift chamber”’

(DC), situated in a solenoidal magnet with a

nominal field strength of 4.5 kG, and a six layer cylindrical (TC),

located b e t ween the DC and the beam pipe.

grouped together

in twelve “superlayers”

trigger drift chamber[”

The layers of the DC are

each consisting of six layers of wires.

The wires in the odd numbered superlayers are parallel to the cylindrical and the wires in alternating

axis

even numbered superlayers are pitched at +66 mrad

and -66 mrad with respect to the cylindrical

axis to provide stereo information.

The DC has an active length of 2.30 m; the number of sense wires per layer ranges from 26 at the inner radius of 25.0 cm to 136 at the outer radius of 144.4 cm.

The TC consists of six layers of axial sense wires in 4 mm radius

aluminized

mylar tubes.

The inner layer of this chamber has 72 wires at a

radius of 9.5 cm, and the outer layer has 112 wires at a radius of 14.8 cm. The average point measurement

resolution

for the TC. The non-vertex-constrained (GeV/c)-l,

is 175 pm for the DC and 90 pm

momentum

resolution is oP/p2 = 0.0034

measured with 14.5 GeV/c electron tracks which traverse all layers

of both chambers (I cos 81 < 0.62) in Bhabha scattering events. 3

The position

resolution

of a track extrapolated

back to the beam position,

in the plane transverse to the beam axis, can be measured from the distribution of separation

distances of collinear

Bhabha tracks.

This resolution,

which is

the standard deviation of the separation distance distribution

divided by 4,

oezt = 78 pm for our detector.

to this resolution

due to multiple

scattering

We calculate the contribution

in the beam pipe and TC chamber inner wall to be

0~s = 152 pm/p for tracks perpendicular The average beam interaction events for luminosity the vertical

(horizontal)

beam-beam

overlap.

to the beam axis.

point was derived from Bhabha-scattering

subsamples of -

0.5 pb-l.

Tracks within

a measurement distributions

f52

mrad of

plane were used to measure the size oZ (or,) of the PEP The distributions

of impact parameters

plane, relative to the derived average interaction true beam-beam

is

in the transverse

point, are convolutions

overlap, the movement of the average interaction period, and the track extrapolation

error.

point within

For our data, these

are well described by Gaussians centered at zero.

beam size measured from these distributions,

of the

The effective

with the track extrapolation

error

unfolded, is cr, = 403 f 14 pm and cry = 52 f 11 pm. An event was included in the multihadron

data sample used in this analysis if

it had at least five charged tracks, a reconstructed

primary vertex with a position

less than 4 cm in radius and 10 cm in z (distance along the beam axis) from the beam intersection

point, more than 25% of the beam energy in charged tracks,

and more than 50% of the beam energy in charged tracks and neutral clusters in the calorimeters which surround the DC.““’

The charged tracks used to meet the

above cuts were each required to have at least 0.1 GeV/c of momentum transverse to the beam, an angle 8 with respect to the beam such that 1cos 81 < 0.8, and a distance of closest approach to the beam intersection radius and 10 cm in z. In addition,

point of less than 6 cm in

if a charged track had a momentum

than 12.5% of the beam energy, its contribution to this amount. 4

greater

to the energy sums was limited

The method of tagging D*+ + ?r+D” decays”’ pion has been in use for many years,“] divided each multihadron

with the spectator charged

and is well established.

In our analysis, we

event into two hemispheres using a plane perpendicular

to the thrust axis determined from charged tracks, and tried every charged track with p > 0.4 GeV/ c as a spectator pion in association with the Do candidates in the same thrust hemisphere. track combination two-track

with a K-z+

combination

combination

A Do candidate was chosen as either: I) a twomass between 1810 and 1920 MeV/c2,

with 1580 < mK-*+

with 1810 < mK-*+r-r+

< 1700 MeV/c2,

< 1920 MeV/c2.

II) a

or III) a four-track

No particle identification

We tried all possible K and 7~ mass assignments,

was used in our analysis.

consistent with the requirement

that the kaon has a charge opposite that of the

spectator pion, and that the total charge of the Do candidate be zero. Cases I and III correspond to fully reconstructed

Do decays, and Case II corresponds to

the satellite resonance of the decay Do -+ K-r+?rO, is reconstructed. of the K-T+

where only the K-T+

Due to the dynamics of the Do -+ K-r+n”

invariant

mass distribution

mass

decay, a large part

is in the mass region of Case II.“’

We

required each track from the Do candidate to have p > 1.2 GeV/c for Case II and p > 0.8 GeV/c for Case III. For Case I we required that the helicity angle of the K-

in the Do rest frame satisfy -0.8

equivalent to a 1.0 GeV/c minimum Do decay. In addition,

< cos0H < +0.8, which is nearly

momentum

cut on the K-

we required that x, the combined energy of the spectator

pion and the Do candidate

divided by the beam energy, be greater than 0.5.

This was done because the D*+ signal has less combinatorial smaller contribution

and r+ from the

from the decay B --) D*+X

from B meson decay complicate negligible B meson lifetime. Do candidates in a thrust

and a

at x > 0.5.“’ The D*+ mesons

the Do lifetime measurement

due to the non-

For Case III there were occasionally more than one hemisphere which satisfied all the above cuts. Since

these duplicate combinations cases we chose the K-n+rlr-r+

background

generally shared many of the same tracks, in these combination

with mass closest to the Do mass as

the only Do candidate in that thrust hemisphere. 5

We then calculated the difference in mass (Am) between the spectator pionDo candidate pair and Do candidate. I and III)

are histogrammed

together

The fully reconstructed in Fig.

Do decays (Cases

la, and show a narrow signal,

consistent with our expected resolution, centered near the world average valuetl” of Am = mg+ produced

-

mD0

= 145.45 MeV/c2.

satellite decay

a broad Am enhancement between 140 and 152 MeV/c2

The degraded Am resolution momentum

(Fig.

in Case II is caused by the uncertainty

lb).

in the Do

due to the missing TO. We took our D*+ signal for Cases I and III

to be all spectator MeV/c2

The Do ---) K-rr+ro

pion-Do candidate combinations

and, for Case II, those combinations

with 143.5 < Am < 147.5

with Am < 152 MeV/c2.

The Do

mesons used for our lifetime measurement were the decay products of these D*+ mesons. In addition to the cuts described above, we required that at least two tracks from the Do decay contain four or more TC position measurements (hits), and we used only these tracks in the secondary vertex fit to determine the Do decay point and its associated error matrix.

For the Do + K-?r+?r-?r+

decay, this meant

that not all tracks from the Do decay were required in the vertex fit, though the full information

from the Do decay was used in all other stages of the analysis.

This was done because it was possible for some tracks from the Do decay not to have at least four good TC hits because of track overlap, but by the time they traversed the DC, to be cleanly separated and reconstructed, angular resolution

and multi-hit

capability

due to the superior

of the DC.

There were 53 Do candidates which passed the above requirements a vertex fit probability candidates

greater than 0.01. We estimate that 17 f 7% of these

are from random combinatorial

B ---) D*+X

and had

background,

and 6 f 3% are from

decays. We scanned the events to make sure that none of the tracks

included in the Do vertex fit had erroneous or ambiguous TC hits associated with them, and to inspect the rest of the event for obvious event selection or track reconstruction

errors which would affect the lifetime measurement.

Do candidates failed these scanning criteria. 6

None of the

The formulae for calculating the most probable path length (in the transverse plane) of the Do, from its production

point to its decay vertex, and the associated

error, are given in Ref. 2. These formulae are functions of the vertex position, the beam position,

the associated error matrices, and the reconstructed

vector of the Do.

The path length was converted into a lifetime

momentum by dividing

it by the measured ~Pc sineD, where 00 is the angle between the Do direction and the beam axis. KK-r+

For the Do +

K-vr+r”

and rIT+ tracks are reconstructed, combination

to be a negligible

satellite

the momentum

decay, where only the vector and mass of the

were used. Monte Carlo simulations show this approximation source of error.

A histogram

of the distribution

of lifetime

measurements, which has a mean of 0.48 f 0.11 psec, is shown in Figure 2. We performed

a maximum

likelihood

fit to the lifetime measurements.

The

fitting function, which is also described in detail in Ref. 2, was the sum of a decay exponential

convoluted with a Gaussian resolution function for the directly pro-

duced Do mesons in the sample, the same function convoluted with another decay exponential to simulate the B + D”X

background, and a Gaussian with its mean

offset from zero lifetime to describe the background from random combinations of tracks in multihadron

events which pass all of the Do selection requirements.

We used a B meson lifetime of 1.16 psec in the background function.[111 The offset of the Gaussian used for the combinatorial weighted-mean

background was taken from the

lifetime of a side-band sample. This sample consisted of all two-

and four-charged-track

combinations

which passed the same cuts as were applied

to tracks in the Do signal region, except that we required invariant

mass to be between 2000 and 2400 MeV/c2.

the combination’s

To increase the statistics

in this sample, we did not require the total charge of a combination and did not require a spectator pion with which the combination Am cut.

This sample should have approximately

the random combinatorial of the distribution 0.17f0.01

background

to be zero

would pass the

the same mix of lifetimes as

in the D*+ sample. The weighted-mean

of side-band lifetimes is 0.19 f 0.01 psec for the data, and is

psec for a Monte Carlo data sample generated with the Lund JETSET 7

6.3 program1121 and passed though our detector simulation. The fit to the data gave a Do lifetime only). The contributions

of 0.44+::$

psec (statistical

error

to the estimate of the systematic error from the various

parameters used in the fit are shown in Table 1. As in Ref. 2, we included a scale factor in our fits to account for systematic mis-assignment

of track errors. The

scale factor was set to 1.2, based on a comparison of the track error matrices and the measured separation

distance distribution

events. Other checks of track resolution scale factor as a free parameter within

the amount (f0.2)

for tracks in Bhabha scattering

in multihadron

events, and leaving the

in the fit, gave values for the scale factor well

by which we varied it to estimate its contribution

to

the systematic error. The fraction

of random combinations

taking the shape of the Am distribution studies and normalizing la and lb.

in the Do sample was estimated for this background

by

from Monte Carlo

it to the Am regions above the D*+ signals in Figures

The significant

positive lifetime

that there are some non-zero lifetime decays in this background,

of the side-band sample confirms

decay products

from charm and beauty

so we assign a large range of possible values to the

Gaussian offset when estimating

its contribution

to the error. For another check,

we generated Monte Carlo samples with the Do lifetime set to 0.1, 0.4, and 0.8 psec. We analyzed these samples, which had complete detector simulation,

with

the same programs as were used for the real data. The measured lifetimes of the Do candidates in these samples were consistent with the input lifetimes to within ho.03 psec, which we took as the systematic error contribution

from any overall

lifetime measurement offset. While the average B meson lifetime is now reasonably well determined, possible that the B” and Bsignificantly

different.

lifetimes and their branching

ratios to D*+X

it is are

Because of this, we took the range over which we varied rB

in our systematic error estimate to f0.4

psec. The Monte Carlo estimate of the

fraction of Do mesons from B decay is likewise dependent on the fragmentation 8

functions used for c and b quarks, the dynamics of the B + D*+X

decay, and the

above branching ratios. We have adjusted the Monte Carlo parameters to give the good agreement with measured fragmentation and the D*+ momentum uncertainty

functions,

B decay multiplicities,

spectrum from B decays, and estimate the systematic

in the fraction

of Do mesons from B decays in our sample to be

f0.03. We have investigated rotation

other possible sources of systematic

error, such as a

of the TC with respect to the DC, and a difference between the real

beam spot size and the one used ,in the decay length calculation, these to be negligible.

When added in quadrature,

error from the contributions

in Table 1 is f0.06

but we find

the total estimated systematic psec.

In conclusion, we have measured the lifetime of the Do meson to be 0.44fi:$f 0.06 psec. This result is consistent with our previous measurement lifetime and with the current world average,“” measurement

from the Fermilab

0.435 f 0.015 f 0.010 psec.[ls’

which is dominated

Tagged Photon Spectrometer

of the Do by a recent

collaboration

of

REFERENCES

1. Proposal for the Mark II at SLC, CALT-68-1015

(1983).

2. L. Gladney et al, Phys. Rev. D34 (1986) 2601. 3. R.H. Schindler et al, Phys. Rev. D24 (1981) 78. 4. G. Hanson, Nucl. Instr. Meth. A252

(1986) 343.

5. W.T. Ford et al, Nucl. Instr. Meth. A255 6. The charge-conjugate

(1987) 486.

reactions are implicitly

included for all reactions de-

scribed in this paper. 7. G.J. Feldman et al, Phys. Rev. Lett. 38 (1977) 1313. 8. G. Goldhaber,

Proceedings

of the XVIIIth

Rencontre

De Moriond,

La

Plagne (1983). 9. See, for example, S. Bethke, Z. Phys. C29 (1985) 175. 10. Particle Data Group, Phys. Lett.

170B, April 10, 1986.

11. M. G. D. Gilchriese, Proceedings of the XXIII

International

Conference on

High Energy Physics, Berkeley (1986). 12. T. Sjostrand,

Computer

Bengtsson, LUTP

Phys.

Comm.

39 (1986) 347; T. Sjostrand,

86-22 (1986).

13. J.C. Anjos et al, Phys. Rev. Lett. 58 (1987) 311. 10

M.

Source of

Estimated

Contribution

Systematic Error

Range

Resolution scale factor

1.2 f 0.2

f0.03

psec

Fraction of signal from

0.17 f 0.07

f0.02

psec

f0.19

f0.01

psec

f0.03

psec

f0.03

psec

random combinations Offset of Gaussian for random combinations

f0.10

psec

Overall measurement offset Fraction of signal

0.06 f 0.03

from B + D”X 1.16

Average B lifetime for B + D”X Table

1. Contributions

f0.40

ho.01 psec

psec

to estimated systematic error.

11

FIGURE 1. The Am distribution

CAPTIONS

for all Do candidate-spectator

pion combinations,

as

described in the text, where (a) the Do candidate is a fully reconstructed K-K+

or K-rIT+7r-rIT+ decay, and (b) the Do candidate is in the satellite re-

gion, corresponding to the high mass K-n+

reflection of the Do + K-d?r”

decay, 2. The distribution

of measured Do lifetimes.

the signal and background functions, to the fit of random combinatorial contribution

from B --+ D”X

The solid curve is the full fit of

the dotted curve is the contribution

background,

background.

12

and the dashed curve is the

I

50

(a)

40 30 20 IO 0 12 -

(b)

1

8 --

4 --

0 100 4-87

120

140 Am

I

160

(MeV/c2)

Figure

1.

5766A2

V

IO 8

-

6

-

4

-

2 0 -4 4-87

-2

2

0 Lifetime

Figure

(psec)

2.

4 5766A

1