Muon Catalyzed Fusion (µCF) - Recent progress and future plan

K. Ishida (RIKEN)Introductiondd formation:

wider range of temperature, phase and ortho/para ratio of D2 dt formation:

first experiment with T2/ortho-D2 mixtureCF with intense muon beam

in collaboration withK. Nagamine1,2,3, T. Matsuzaki1, N. Kawamura2, H. Imao1, M. Iwasaki1, Y. Matsuda1, S. Nakamura1**, A. Toyoda2, M. Kato4, H. Sugai4, A. Uritani5, H. Harano5, G.H. Eaton6 1RIKEN, 2KEK, 3UC Riverside, 4JAEA, 5AIST, 6RALpresent address *U. Tohoku , **RIKEN

NuFact06＠ UC Irvine 2006.08.28

muon catalyzed fusion (µCF) - principle and motivations

After injection of muons into D/T mixture (or other hydrogen isotopes)

Formation of muonic atoms and muonic moleculesIn small dtµ molecule, Coulomb barrier shrinks and d-t fusion followMuon released after d-t fusion

- muon works as catalyst -

History1947 Hypothesis of µCF (Frank)1957 observation of pdµ fusion (Alvarez)1966 observation of resonant ddµ formation1967 hypothesis of resonant formation Vesman)1979-82 observation of large dtµ formation rate1987 observation of x-rays from (µ)+ (PSI,KEK)1993 large ddµ formation rate in solid1995 study with eV beam of (tµ)1996 systematic study starts at RIKEN-RAL

tµ

dµ

µdtµ

n

µ

nuclearfusion

d14MeVneutron

3Heµ

Sticking

Maximizing µCF efficiency

.MotivationsAtomic physics in small scale

rich in few body physics problemsProspect for applications

(fusion energy, neutron source)muon production cost (~5 GeV) vs fusion output (17.6 MeV x 120)

Observables

(1) Cycling rate c () (vs 0: muon life)

dtµ formation tµ + D2 [(dtµ)dee]

(2) Muon loss per cycle W ()muon sticking to -particle, etc

Fusion neutron disappearance rate : n = 0 +Wc

Number of fusion per muon:

Yn = c/n = 1 / [(0/c)+W] ()

tµ

dµ

µdtµ

n

µ

dtµformation

nuclearfusion

effective stickings=(1-R)s0

17.6MeV x Ynenergy output

injectedmuon

d

K/KX-ray

14MeVneutron

3Heµtransfer toHe

Outstanding problems in CF

Main CF processes are understood, however still discrepancies …1. Muonic atom

acceleration during cascade, hyperfine transition, transfer2.Molecular formation rates

temperature dependence non-linear density (3-body?) effectortho/para

3. Muon to alpha stickinginitial sticking atomic process, muon reactivation

4. Helium effectmuon transfer processhelium in solid D/T

Progresses are being made in each of the problems.Focus is given to molecular formation rate in this talk.

dt formation

transfer

Key process of µCF - dtµ formation

Auger formation : 106 /s (as slow as muon decay rate 0.45 x 106/s)tµ + D2 (dtµ) + D + e-

Resonant molecular formation : 106-109 /stµ + D2 ((dtµ) dee)

(dtµ binding energy ~ excitation of complex molecule)(tµ energy to match the small

energy difference)

Thus,dependence on temperaturedependence on initial states such as tµ spin state, D2 states

We could change (enhance) µCF by controlling initial D2 states.

E = 11dt + 0

t

t + (D2)iKi → [(dt)11dee]*vfKf

t

d d dtd

e-

e-

e-

e-

→＋

dtU

R

D2

E11dt

0t

= 0

= 1

= 2

J, = (0,0)

J,v=(1,1) = 0

[(dt)dee]

~0.3eV

Present understanding of dtµ formationdtµ molecule formation

large formation rate (~109/s) by resonant formation mechanismis established (temperature dependence, etc)still many surprises

non-trivial density dependence even after normalizationthree-body effect : tµ + D2 + D2’ ((dtµ)dee) + D2’’

low temperature & solid state effect

density

How about ddµ ?Same resonant formation process applies for ddµ formation in pure D2

Slower compared to dt-µCF, and with lower energy output per fusionStill, study should be done in parallel with study of dtµ

No need of tritium Simpler cycling process (pure D2)

dµ + D2 [(ddµ)dee] In analogy to dtµ case, temperature dependence dµ Hyperfine effect (F=1/2,3/2) D2 molecule effect

dd formation: understanding before 2000Resonant ddµ formation is nearly established in gas　 fitting by theoretical curve gives

precision determination of shallow state binding energy in ddµ (-1.97 eV) [Petitjean et al]

which is comparable to the valueby precise three-body calculation

However, in liquid and solid, large deviation from the theoretical curve was observed

[Knowles et al, Demin et al]

Theoretical prediction of dd formation rate dependence on D2 ortho-para state.

(ortho)

(ortho)(para)

Ki=0

2

Ki=3 4 2

10 20 300

~5K

~20K~50K

F=3/2

F=1/2

Resonance l i nes and Maxwel l i an di stri buti on

t energy[meV]

New parameter: Ortho- and para-D2 in CF

D2 populates several rotational states (J=0,1,2,3,4,…)With normal D2, the resonance condition for all these are mixed.

Ortho-D2 (dd nuclear spin coupled with 0,2, spin J,v symmetry under dd exchange allows only J=even)Para-D2 (similarly, for coupling with 1, J=odd allowed)Normal mixture is ortho:para=2:1(for H2, para H2 has J=even)

Resonance condition for each state should be separately measured.

First measurement with normal- and ortho-D2 at 3.5KToyoda et al (2000) @ RAL, TRIUMF (fusion proton)

E968/E1061 Experiment @TRIUMF

Measure dd CF in D2 target (Since 2003)ortho/para controlledlow temperature target (5K - 36 K)gas/liquid/solid phaseseveral densities (0.03 to 1.2 of LHD)

detectiong fusion neutrons

TRIUMF M9B muon channel decay - beam from 500 MeV x 200 A proton cw beam was necessary for dd fusion 1) good time resolution (<1 ns) 2) suppression of muon capture neutrons background by coincidence with delayed me-decay 10 s beam gate limits the beam rate (50 k/s) 50 MeV/c muon beam (cryogenic Cu target cell, 1.5 Mbar)

Experimental setup @TRIUMF (E968/E1061)D2 (ortho, normal or para) preparation

ortho-para ratio analysis with Raman spectroscopyTarget cell (30cc liquid/solid, or 200cc pressurized gas)muon beam counters B1+B2µe-decay electron counters E1-E8 (>50% solid angle)neutron detectors(NE213) N1-N4

0 10cm

target

2006 RUN

2004

muonbeam

gas preparation

D2 target and detectors

Ortho/para-D2 preparation and analysis

Ortho D2 (established)

ortho-para conversion with Al2O3/Cr2O3 catalysis

Para D2 (some success in Jun06 RUN)

preferential adsorption on Al2O3 at ~20K

99% para D2 had been reported [D.A. Depatue and R.L. Mills, 1968]

our case 55% para large gas quantity (30 liter), temperature control etc

Analysis by laser Raman spectroscopy

normal(o:p=0.67:0.33)

para-rich(o:p=0.45:0.55)

ortho(o:p=1:0)

(J=02)(24)

(46)

(13)(35)

Fusion neutron time spectra: gas D2 (36K)

Fusion from resonant dd formation (=0.17, 36K) increased for ortho-rich D2

However, some structure in very early timing.

Understanding non-exponential time structure

Resonance calculation [by Adamczak]

with time evolution of d energy

1

10

100

600ns

Imao’s thesis

deceleration passing through dip in resonance spectrum

RUN in Jun 2006

Gas data at =0.03 (28K and 36K), 0.07(36K), 0.17(36K)were recently obtained for normal-, ortho-, and para-rich(55%) D2

to see temperature effect, density effect, non-exponential e\ffect (para)

analysis is still in progress (signal selection, b.g. subtraction etc)=0.07 =0.03

Typical fusion neutron time spectra: liquid D2

fusion from resonantly formed ddµ (fast component) decreased in ortho-D2 for all the temperature data (5K-36K) for solid and liquid D2

non-resonant

resonant

SUM of 19K-23K

Jun06 New data(Liquid)

Gas

dd summaryData for gas D2 (=0.17, 35K) was consistent with calculations based

on idealistic gas modelNon-exponential structure of fusion neutron time spectra is qualitativel

y understood by slow thermalization

In liquid and solid phases, the observed ortho-para dependence of dd formation rate was opposite to the prediction based on a simple gas model in all the temperature range measured (3K - 36K)

Target D2 density (rather than temperature) is responsible for the reversal

Theory is being developed to include density and phase effectresonance energy shift, broadening

dt case: theoretical calculations1. Idealistic gas model tµ + D2 [(dtµ)dee]

low temp. resonance only for para-D2

2. with Condensed matter effect under development [Adamczak]

Even larger effect of ortho-para D2 expected!

[Faifman]

23K liquid

[Adamczak]

Expected effect in D/T(50%) mixture

Prepare ortho-D2 or para-D2 and mix with T2 Very high cycling rate could be expected(>2 of normal rate) followed by decay due to1. molecular equilibration process

D2 + T2 　 2DT (~68 hour in D/T(50%))tµ + D2 [(dtµ)dee]tµ + DT [(dtµ)tee] dtµ

0,D2>>dtµ0,DT)

2. ortho-para equilibration by radiation effecto-p conversion by paramagnetic T atom~16 hours in D/T(50%) at 14K

Cycling Ratec = 1/(d+t)=1/[ q1sCd/(dtCt) + (3/4)/(t

1,0Ct)+ 1/(

dt-D2(o,p)CD2(o,p) + dt

0,DT CDT)]

full eq.

av. time waiting d->t transfer

av. time in t(F=1) state

av. time before dt formation

c

Time [hr]

µCF at RIKEN-RAL Muon Facility

RIKEN-RAL Muon at ISIS Intense pulsed muon beam

(70ns width, 50 Hz) 800MeV x 200µA proton 20~150MeV/c µ+/µ- muon5x104 µ-/s (55MeV/c)

and tritium handling facility

µCF experiment

Proton beam line

µSR

Slow µ (Matsuda)

µA*

µCF target and detectors

Cryogenic target : 1 c.c. liquid or solid D-TDetectors with calibration

fusion neutrons, X-rays, µe decay120 muon stops per pulse106 fusions/s

0 100mm

Muon

90 400

Neutron detectors

µe counters

D-T Target

Si(Li) X-ray detector

Be Windows

840

Superconducting magnet

~

~

~

~

~

~ n

X-ray

e

muonµCF setup

(RIKEN-RAL Port1)

Preparation of ortho-D2+T2 target

1) Production of ortho D2, analysis of ortho/para ratio

2) Evacuation of TGHS3) Charge ortho D2 into target through TGHS and solidify4) Extraction of T2 from getter and solidify into target5) mix D2 and T2 by melting, start CF measurement

6) 　 after several days, turn to gas to force equilibration,restart CF with equilibrated gas

Preliminary result of the first RUN: Cycling rate ： c

D2+T2

D2+T2+DT(eq)

the difference between normal D2 and orthoD2 was small (~5%)

forced equilibrationat gas state

（ Sep 2005)

(Aug 2001)

Result: Muon loss probability W

D2+T2+DT(eq)

D2+T2

Inpurities removes muon even at ppm levelThe impurity disappeared in liq D/T in a few hours, by self condensation(?) impurity

effect

largest contribution is from -to- sticking, butas dt formation slows downloss through side path dd, tt formation increases

Why the ortho-D2 effect was not clearly seen?

Possible reasons1) fast ortho-para conversion (?)

in the D2 inlet path, in D2/T2 liquid masked by impurity effect (first few hours)2) ortho-para effect is small (?)

non-thermalizationchange of resonance condition in condensed matter, similarly to dd case

Near future plansmeasurement with reduced impurity and para D2study of ortho-para conversion rate at tritium lab. in JAEAin situ Raman analysis with new D/T target with optical window

Other ongoing studies of µCFUnderstanding mechanism and parameters of key processes1. Muon-to-alpha sticking loss [my talk in NuFact04]

muon-to-alpha sticking x-rays

2. Muon loss to accumulated 3He from tritium decay [my talk in NuFact04]tHe molecule formation and decay

3. dt formaiton in wider and exotic target conditions low temperature solid D/T, high density D/T, ortho-para D2

4. fusion in tt particle correlations in 4He-n-n exit channel

Use of intense muon beam in CFIntense muon beams will definitely contribute to the study of µCF

1) efficient search of more and more target conditions

2) short-lived extreme conditions (laser, plasma etc)

3) better S/N

4) exotic beams from µCF

5) intense neutron source

Summary

A clear effect of ortho-para ratio was observed for ddµ formation in D2

In gas case, effect was consistent with theoretical prediction.(need to include thermalization process)In the liquid/solid case, the effect was opposite,possible modification of resonance condition by density effect

Even larger effect was predicted for dtµ formationThis opens up possibility for enhancement of µCFThe first measurement failed to give conclusive result. Further study is planned.

There are also other ongoing studies on CF.Intense muon beam is indispensable for the study of CF.