diamond bremsstrahlung radiator assessment using x-ray topography at chess r.t. jones 1, i....

1
Diamond Bremsstrahlung Radiator Assessment using X-ray Topography at CHESS R.T. Jones 1 , I. Senderovich 1 , K. Finkelstein 2 , F. Klein 3 , and P. Nadel-Turonski 3 Diamond Requirements Diamond Requirements Minimum size: 5 mm x 5 mm 5 mm x 5 mm Orientation: [100] [100] Orientation error: 5° maximum 5° maximum Mosaic spread: 20 μr r.m.s. maximum 20 μr r.m.s. maximum (integrated over the whole crystal) Thickness: 20 μm maximum 20 μm maximum Variation in thickness: ±5 μm ±5 μm maximum maximum Abstract Rocking curve topographs of large diamond monocrystals were taken using the CHESS C-line. These diamonds are needed for use as bremsstrahlung radiators to produce 9 GeV polarized photons for the GlueX experiment at Jefferson Lab. The requirements for the diamond radiators for GlueX, in terms of their very small whole-crystal mosaic spread and their small thickness, together with the restrictions on how they can be mounted within from the intense electron beam environment in which they must operate, demand significant advances over what is established practice at existing coherent bremsstrahlung facilities. Up to the present, such facilities have relied on well-established source of high- quality diamonds of sufficient area (natural gems, HPHT synthetics) but advances in diamond monocrystal growth using chemical-vapor deposition (CVD) have been reported by the leading vendor Element Six. The GlueX group obtained one of these so-called type-III diamonds on loan from the Brookhaven National Laboratory Instrumentation group. This sample was mounted using established techniques and rocking curve measurements were taken. Initial results showed rocking curve peak widths much larger than that expected for perfect diamond, but investigation showed that this was caused by micro-mechanical vibrations in the mount. Once these vibrations were eliminated, near-perfect rocking curve peaks were observed for the entire area of the sample. Extracting mosaic spread Extracting mosaic spread from X-ray rocking curve from X-ray rocking curve peak widths peak widths = 2 d sin() instrumental sources 1.energy spread of beam 2.divergence of beam crystal sources 1.natural width + impurities 2.mosaic spread 3.micro-mechanical motion d large area, highly parallel X-ray beam from C-line large area, highly parallel X-ray beam from C-line monochromator monochromator X-ray CCD camera diamond crystal Goals for the May 2009 run: Goals for the May 2009 run: 1.Measure mosaic spread of new type-III CVD diamond monocrystal from Element Six. 2.Investigate contribution to rocking curve width stemming from mechanical mounting effects. 3.Measure the effects of radiation damage on a used diamond radiator from Jefferson Lab. iversity of Connecticut, Storrs, CT rnell University High Energy Synchrotron Source, Ithaca, NY tholic University of America, Washington DC Diamond mechanical mount Diamond mechanical mount 1.Must have “zero” material within ± 5 mm around crystal 2.Concept: massive “picture frame” + fine wires + glue 20 micron ø W-Au wire stretched to 2.5 g tension optical microscope image Sample CVD type-III diamond monocrystal manufactured by Element Six, on loan from Brookhaven National Lab Instrumentation Group 4mm G 4mm G 300m CAD rendering of mount frame First look at CVD diamond: (2 2 0) reflection First look at CVD diamond: (2 2 0) reflection 340 340.1 340.2 340.3 340.4 340.5 340.6 0 0.2 0.4 0.6 0.8 1 rocking angle (m rad) Intensity Intensity (arb. units) 200 r (actual) Possible explanations considered: broad energy spectrum or large divergence from the monochromator. large mosaic spread in the diamond. large mechanical strain in the diamond (e.g. thermal, bending strains). micro-mechanical motion of the sample. 8 r (expected) ? ? not considered in the past, but suggested by rocking curve shown ! Consider: a simple oscillator model Consider: a simple oscillator model t peaks appear where instantaneous velocity is zero ≡ turning poin ± 100 r ± 200 nm motion at crystal edges: too small to see by eye! Possible sources of vibrations: 1. QM zero point motion (just for fun) 6.4 x 10 -14 radians 2. Boltzmann thermal motion 1.7 x 10 -8 radians 3. Ambient goniometer vibration spectrum unknown 4. Ambient vibration resonance-enhanced fundamental fundamental ~200 Hz ~200 Hz 343 343.2 343.4 343.6 343.8 344 344.2 0 0.2 0.4 0.6 0.8 1 rocking angle (m rad) Intensity Intensity (arb. units) Loading the wires with extra mass shifts down the frequencies of the fundamental resonances. Test of #4: shift the mechanical resonance of the mount, rescan the rocking 800 r a factor 4 worse! 375.1 375.15 375.2 375.25 375.3 0 0.2 0.4 0.6 0.8 1 rocking angle (m rad) Intensity Test of #3: decouple/damp goniometer vibration using a “soft mount” (wax) Intensity (arb. units) Diamond is pressed into a lump of soft wax and mounted on a stiff post. 10 r FWHM Conclusions Rocking curve measurements show that the intrinsic mosaic spread of the new generation of type-III diamonds from Element Six meets the requirements for the GlueX experiment. Mounting a diamond radiator on thin wires attached to a massive frame, as is the current practice at Jefferson Lab and is subject to micro-mechanical vibrations that completely obscure the natural rocking curve width of the diamond, leading to rocking curve widths that are an order of magnitude larger than the GlueX requirements. Thermal vibrations alone are not sufficient to explain these large oscillations. They are not due to an accidental resonance of the diamond mount with some noise source present in the area where the measurements were carried out. Once a way is found to mount the diamond that eliminates or damps the oscillations to a negligible level, further studies will be necessary to show that narrow rocking curves This work is supported by the United States Department of Energy through the Thomas Jefferson National Accelerator Facility, and by the National Science Foundation through grant no. DMR0225180. Intensity (arb. units) (r) (2 2 0) scattering intensity @ 15 keV fit to sum of 3 Gaussians + unif. bg. individual Gaussians from fit ideal lineshape for diamond (2 2 0) ideal FWHM = 5r + 6r natural instrumental

Upload: cameron-goodwin

Post on 03-Jan-2016

215 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Diamond Bremsstrahlung Radiator Assessment using X-ray Topography at CHESS R.T. Jones 1, I. Senderovich 1, K. Finkelstein 2, F. Klein 3, and P. Nadel-Turonski

Diamond Bremsstrahlung Radiator Assessment using X-ray Topography at CHESSR.T. Jones 1, I. Senderovich 1, K. Finkelstein 2, F. Klein 3, and P. Nadel-Turonski 3

Diamond RequirementsDiamond Requirements

Minimum size: 5 mm x 5 mm 5 mm x 5 mm

Orientation: [100] [100]

Orientation error: 5° maximum5° maximum

Mosaic spread: 20 μr r.m.s. maximum20 μr r.m.s. maximum (integrated over the whole crystal)

Thickness: 20 μm maximum20 μm maximum

Variation in thickness: ±5 μm maximum±5 μm maximum

AbstractRocking curve topographs of large diamond monocrystals were taken using the CHESS C-line. These diamonds are needed for use as bremsstrahlung radiators to produce 9 GeV polarized photons for the GlueX experiment at Jefferson Lab. The requirements for the diamond radiators for GlueX, in terms of their very small whole-crystal mosaic spread and their small thickness, together with the restrictions on how they can be mounted within from the intense electron beam environment in which they must operate, demand significant advances over what is established practice at existing coherent bremsstrahlung facilities. Up to the present, such facilities have relied on well-established source of high- quality diamonds of sufficient area (natural gems, HPHT synthetics) but advances in diamond monocrystal growth using chemical-vapor deposition (CVD) have been reported by the leading vendor Element Six. The GlueX group obtained one of these so-called type-III diamonds on loan from the Brookhaven National Laboratory Instrumentation group. This sample was mounted using established techniques and rocking curve measurements were taken. Initial results showed rocking curve peak widths much larger than that expected for perfect diamond, but investigation showed that this was caused by micro-mechanical vibrations in the mount. Once these vibrations were eliminated, near-perfect rocking curve peaks were observed for the entire area of the sample.

Extracting mosaic spread from Extracting mosaic spread from X-ray rocking curve peak widthsX-ray rocking curve peak widths

= 2 d sin()

instrumental sources

1.energy spread of beam

2.divergence of beam

crystal sources

1.natural width + impurities

2.mosaic spread

3.micro-mechanical motion

d

large area, highly parallel X-ray beam from C-line monochromatorlarge area, highly parallel X-ray beam from C-line monochromator

X-ray CCDcamera

diamond crystal

Goals for the May 2009 run:Goals for the May 2009 run:1. Measure mosaic spread of new type-III CVD diamond monocrystal from

Element Six.2. Investigate contribution to rocking curve width stemming from mechanical

mounting effects.3. Measure the effects of radiation damage on a used diamond radiator from

Jefferson Lab.

1 University of Connecticut, Storrs, CT2 Cornell University High Energy Synchrotron Source, Ithaca, NY3 Catholic University of America, Washington DC

Diamond mechanical mountDiamond mechanical mount1. Must have “zero” material within ± 5 mm around crystal2. Concept: massive “picture frame” + fine wires + glue

20 micron ø W-Au wirestretched to 2.5 g tension

optical microscope image

Sample CVD type-III diamond

monocrystal manufactured by Element Six, on loan from Brookhaven National Lab Instrumentation Group

4mm G 4mm G 300m

CAD rendering of mount frame

First look at CVD diamond: (2 2 0) reflectionFirst look at CVD diamond: (2 2 0) reflection

340 340.1 340.2 340.3 340.4 340.5 340.60

0.2

0.4

0.6

0.8

1

rocking angle (mrad)

Inte

nsity

Inte

nsi

ty (

arb

. u

nit

s)

200 r (actual)

Possible explanations considered: broad energy spectrum or large

divergence from the monochromator. large mosaic spread in the diamond. large mechanical strain in the

diamond (e.g. thermal, bending strains).

micro-mechanical motion of the sample.

8 r (expected)??

not considered in the past, butsuggested by rocking curve shown !

Consider: a simple oscillator modelConsider: a simple oscillator model

t

peaks appear where instantaneous velocity is zero ≡ turning points

± 100 r ± 200 nm motion at crystal edges: too small to see by eye!

Possible sources of vibrations:1. QM zero point motion (just for fun) 6.4 x 10-14 radians

2. Boltzmann thermal motion 1.7 x 10-8 radians

3. Ambient goniometer vibration spectrum unknown

4. Ambient vibration resonance-enhanced fundamental ~200 fundamental ~200 HzHz

343 343.2 343.4 343.6 343.8 344 344.20

0.2

0.4

0.6

0.8

1

rocking angle (mrad)

Inte

nsity

Inte

nsi

ty (

arb

. uni

ts)

Loading the wires with extra mass shifts down the frequencies of the fundamental resonances.

Test of #4: shift the mechanical resonance of the mount, rescan the rocking curve

800 r

a factor 4 worse!

375.1 375.15 375.2 375.25 375.30

0.2

0.4

0.6

0.8

1

rocking angle (mrad)

Inte

nsity

Test of #3: decouple/damp goniometer vibration using a “soft mount” (wax)

Inte

nsi

ty (

arb

. uni

ts)

Diamond is pressed into a lump of soft wax and mounted on a stiff post.

10 r FWHM

ConclusionsRocking curve measurements show that the intrinsic mosaic spread of the new generation of type-III diamonds from Element Six meets the requirements for the GlueX experiment. Mounting a diamond radiator on thin wires attached to a massive frame, as is the current practice at Jefferson Lab and is subject to micro-mechanical vibrations that completely obscure the natural rocking curve width of the diamond, leading to rocking curve widths that are an order of magnitude larger than the GlueX requirements. Thermal vibrations alone are not sufficient to explain these large oscillations. They are not due to an accidental resonance of the diamond mount with some noise source present in the area where the measurements were carried out. Once a way is found to mount the diamond that eliminates or damps the oscillations to a negligible level, further studies will be necessary to show that narrow rocking curves are possible with much thinner diamond samples. This work is supported by the United States Department of Energy through the Thomas Jefferson National Accelerator Facility,

and by the National Science Foundation through grant no. DMR0225180.

Inte

nsi

ty (

arb

. u

nit

s)

(r)

(2 2 0) scattering intensity @ 15 keVfit to sum of 3 Gaussians + unif. bg.individual Gaussians from fitideal lineshape for diamond (2 2 0)

ideal FWHM = 5r + 6r

natural instrumental