Measurement Of Energy Resolution For TES Microcalorimeter With Optical Pulses
Shuo Zhang, Qing-Ya Zhang, CongZhan Liu, Jianshe Liu, Wenhui Dong, Wei Chen
CChinese Physics C Vol. xx, No. x (201x) xxxxxx
Measurement Of Energy Resolution For TES Microcalorimeter WithOptical Pulses *Shuo Zhang
Qing-Ya Zhang , Jianshe Liu , CongZhan Liu Wenhui Dong , Wei Chen , Tsinghua National Laboratory for Information Science and Technology, Tsinghua University, Beijing 100084, China Institute of Microelectronics, Department of Micro/Nanoelectronics, Tsinghua University, Beijing 100084, China Key Laboratory of Particle Astrophysics, Institute of High Energy Physics, Chinese Academy Of Sciences, Beijing, 100049, China
Abstract:
Energy resolution is an important figure of merit for TES microcalorimeter. We propose a laser systemto measure the energy resolution of TES microcalorimeter with a 1550 nm laser source. Compared to method thatcharacterizes the performance by irradiating the detector using X-ray photons from a radioactive source placed insidethe refrigerator, our system is safer and more convenient. The feasibility of this system has been demonstrated inthe measurement of an Al/Ti bilayer TES microcalorimeter. In this experiment, the tested detector showed a energyresolution of 72 eV in the energy range from 0.2 keV to 0.9 keV.
Key words:
TES , Microcalorimeter , Energy Resolution , Laser Pulse
PACS:
TES (Transition Edge Sensor) microcalorimeter, asa very sensitive temperature sensor, has been widelyused in radiation detection. Compared to the traditionalsemi-conductor detectors, TES microcalorimeter has ex-cellent detection efficiency and much higher energy res-olution. TES Microcalorimeter has been playing an im-portant role in the field of X-ray astronomy [1–4], nuclearenergy spectrum[5], X-ray spectroscopy analysis[6], ele-mental analysis[6], neutrino mass measurement[7], darkmatter detection and so on.Generally, the energy resolution of TES mi-crocalorimeter is measured by radioactive sources whoseenergy are limited and discrete. If we need to measurethe energy resolution at other energy, we should inter-rupt the operation of the refrigerator and change theradioactive sources. This process take a lot of time andwill change the test environment.Microcalorimeter detects X-ray base on the thermalsignals, so we can use a thermal source replace radioac-tive source. We use a 1550 nm laser source to replacethe radioactive source as the energy impulse for the testof TES. In comparing with the radioactive source, theequivalent energy and the frequency of laser pulse is ad-justable. We can therefore use the laser to find the proper work point of the TES and get the energy resolution, lin-ear dynamic range, characteristic shape of signal pulsevery quickly. In this paper, we will introduce the sys-tem, and demonstrate the feasibility of this system bymeasuring the energy resolution of an experimental TESmicrocalorimeter we fabricated.
As shown in Figure 1, a 1550 nm laser source with aconstant output power P is used to mimic the TES mi-crocalorimeter, the pulse width t and frequency f mod-ulated by a function generator. Current change signal δI is converted to a voltage signal by SQUID amplifier,after filtered by pre-amplifier and Liner amplifier the sig-nal is analysed by a Multi-channel analyser, then we gotthe energy spectrum of the laser pulse.The bias voltage for the TES is provided by a roomtemperature current source I b and a 8.1 mΩ resistor ( R b )at 2.5 K. The TES branch has a parasitic resistance R p =44mΩ from the wirings. The input inductance of SQUIDamplifier ( L in ) is about 600 nH. Received xx xx 2016 ∗ CAS Strategic Priority A(Grant No. XDA040762)1) E-mail: [email protected]) E-mail: [email protected] c (cid:13) a r X i v : . [ phy s i c s . i n s - d e t ] D ec hinese Physics C Vol. xx, No. x (201x) xxxxxx R b TESR p I b V out R fb L fb SQUID Amp.
Laser sourceModulatorAttenuator
Function generator l = 1550 nm Pulse, f = 50 Hz
Preamp mK stage2.5 KRoomtemperatue Opticalfiber
Oscilloscope
PreamplifierLineramplifierMultchannel
Fig. 1. The diagram of TES testing system usingoptical laser pulses.
As shown in Eq (1), keeping P constantwe can ad-just photon number n t and average energy of laser pulse E ( n t ) by controlling t , here E λ is energy of single photon,for laser with wavelength of 1550nm, E λ = 0 . eV . E ( n t ) = n t ∗ E λ = k ∗ P ∗ t (1)Compare to radiation sources, energy of laser pulseis not constant. Full width of half hight F W HM n t from Multi-channel analyser have two componentsfluctu-ations caused by input photon number δ λ and fluctuationcaused by energy resolution of TES microcalorimeter it-self δ r . F W HM n t = 2 . ∗ δ n t = 2 . ∗ (cid:112) n t ∗ δ λ + δ r (2)Although it is not constant, photon number in a laserpulse obey Poisson distribution π ( n t ), since the energyof laser photon is pretty small compare to X-ray source, n t should be a very big numberso we can simplify it toa Gauss distribution N ( n t , n t ∗ δ λ )where δ λ is a constantvery near 1As a thermal based signal detector, energy resolu-tion of TES is almost constant F W HM r = 2 . ∗ δ r =2 . ∗ (cid:113) k B T C/α I (cid:112) n/ δ r is not changed with n t .So a = δ λ and b = δ r in Eq (2) is constantand we cansimplify Eq (2): δ n t = ( F W HM n t / . = a ∗ n t + b (3)According to Eq (3)we can measure δ n t at different n t linear fitting itsquare root the intercept and multiply2.35, then we get the energy resolution of TES. The TES resistance as a function of temperature,called RT curve is shown in Fig 2. From Fig 2 we can getwidth of RT curve is about 3mKtransition temperatureis about 542mK.
Fig. 2. Ti/Al TES RT curve. n t We measured pulse hightes at different n t and theyshow excellent linearitythat’s mean TES is working ata linear work point of. By pulse hight and shape, wecan get the TES current and resistance. Then we cancalculate the energy of each pulse by integral the powerchange and time.
200 400 600 800 1000 120020030040050060070080090010001100 ene r g y ( e v ) pulse width (ns)
200 400 600 800 1000 1200 2004006008001000
Fig. 3. pulse hight at different n t We measured energy spectrum under different n t wecan see the F W HM n t of the spectrum increase as n t increasing, agree with the theory prediction. c oun t energy(eV) photon number Fig. 4. energy spectrum under different n t n t By analysing the energy spectrumwe can get
F W HM n t under different n t , the result is shown inFig 5, Eq (3) agree with the result very well. We get δ λ = 0 . ± .
022 agree with the theory prediction δ λ = 1. δ r = 38 . ± .
488 corresponding to a δE = 72 . ± . eV energy resolution.
200 400 600 800 1000 120018002000220024002600 s qua r eo f pho t onnu m be r e rr o r photon number
200 300 400 500 600 700 800 900 64006800720076008000840088009200 s qua r eo f ene r g y r e s e l u t i on energy(eV) Fig. 5. δn t under different n t The experiment results agree very well with the the-ory predictions, but 72eV is not a very good energy res-olution. The main reason is as the following description:First energy resolution is affected wiby temperature,
F W HM r = 2 . ∗ δ r = 2 . ∗ (cid:113) k B T C/α I (cid:112) n/
2. OurHe3 refrigerator can only work on temperature higherthan 400mK. If we can reach lower temperature, we mayget better resolution.Secondly, we use Multichannel Signal Analyzer(MSA) to get the energy spectrum directly, this is notthe best method to get the energy resolution, so in thefuture, we need design a new data acquisition system toimprove the energy resolution.
Fangjun Lu give a lot of instruction in the statisticalmethod, We acknowledge the support of CAS Strate-gic Priority A(Grant No. XDA040762), State Key De-velopment Program for Basic Research of China(GrantNo.2011CBA00304) and Tsinghua University InitiativeScientific Research Program(Grant No.20131089314)
References2