Christian Joram / CERN

Transcription

1 Christian Joram / CERN L3-1

2 Outline Lecture 1 Interaction of charged particles Lecture 2 Gaseous and solid state tracking detectors Lecture 3 Calorimetry, scintillation and photodetection Calorimetry electromagnetic cascades hadronic interactions neutrons and neutrinos hadronic cascades Scintillation Photodetection L3-2

3 de/dx de/dx Summary: basic electromagnetic interactions e + / e - Ionisation g Photoelectric effect s bg Bremsstrahlung E Compton effect s E E Pair production s E L3-3

4 Electromagnetic cascades (showers) Electron shower in a cloud chamber with lead absorbers Simple qualitative model Pb plates Consider only Bremsstrahlung and (symmetric) pair production. Assume: X 0 ~ l pair N t t ( t) 2 E( t) / particle E0 2 g e + e - Shower can be initiated by photon OR by electron. Process continues until E(t)<E c tmax total t ( tmax 1) tmax 0 N t max t 0 ln E 0 ln 2 E c After t = t max the dominating processes are ionization, Compton effect and photo effect absorption of energy. E E c L3-4

5 longitudinal energy deposition (a.u.) transverse containment 90% (R M ) Electromagnetic calorimeters A detector, which measures the energy of a e +/- or a high energy g by fully absorbing it destructive method! Longitudinal profile de dt t e t 6 GeV/c e - Shower maximum at t max E ln 0 E c 1 ln 2 95% containment t95% tmax 0.08Z 9.6 Size of a calorimeter grows only logarithmically with E 0 Transverse shower development: 95% of the shower cone is located in a cylinder with radius 2 R M 21 MeV 2 RM X 0 [ g / cm ] Molière E radius c Example: E 0 = 100 GeV in lead glass E c =11.8 MeV t max 13, t 95% 23 X 0 2 cm, R M = 1.8 X cm X 0 46 cm (C. Fabjan, T. Ludlam, CERN-EP/82-37) 8 cm L3-5

6 Energy resolution of a calorimeter General expression Also true for hadronic showers (see below) s ( E) E stochastic term a E b constant term c E inhomogenities bad cell intercalibration non-linearities Also spatial and angular resolution scale like 1/ E noise term Electronic noise radioactivity pile up Homogeneous calorimeter Absorber = active material. Simulation! High E-resolution, no longitudinal shower information. High cost. Quality factor! Sampling calorimeter Absorber and active detector layers. Extra sampling fluctuations s ( E) d E E Shower image Lower cost. L3-6

7 Nuclear Interactions The interaction of energetic hadrons (charged or neutral) with matter is dominated by inelastic nuclear processes. p, p, K, n Excitation and finally break-up of nuclei nuclear fragments (radioactive) + production of secondary particles. multiplicity ln(e) p t 0.35 GeV/c For high energies (>1 GeV) the cross-sections depend only little on the energy and on the type of the incident particle (p, p, K ). sinel s0a s 0 35 mb In analogy to X 0 a hadronic absorption length can be defined A la A because sinel s 0A N s A A inel A 1 3 li A N s total 0.7 similarly a hadronic interaction length l l I a L3-7

8 X 0, l I [cm] Interaction of charged particles Material Z A [g/cm 3 ] X 0 [g/cm 2 ] l I [g/cm 2 ] Hydrogen (gas) (g/l) Helium (gas) (g/l) Beryllium Carbon Nitrogen (gas) (g/l) Oxygen (gas) (g/l) Aluminium Silicon Iron Copper Tungsten Lead Uranium For Z > 6: l I > X 0 l I 1 X 0 l I and X 0 in cm L Z

9 Interaction of neutrons and neutrinos Interaction of neutrons Neutrons have no charge, i.e. their interaction is based only on strong (and weak) nuclear force. To detect neutrons, we have to create charged particles. Use neutron conversion and elastic reactions n + 6 Li + 3 H n + 10 B + 7 Li n + 3 He p + 3 H n + p n + p E n < 20 MeV E n < 1 GeV (H. Neuert, Kernphysikalische Messverfahren, G. Braun Verlag, 1966) n + p n + p E n 100 MeV 100 kev 10-1 ev 10-2 ev 10-3 ev high energy fast epithermal thermal cold ultracold In addition there are - neutron induced fission E n E th 1/40 ev - inelastic reactions E n > 1 GeV L3-9

10 Interaction of neutrons and neutrinos Interaction of neutrinos Neutrinos interact only weakly tiny cross-sections. For their detection we need again first a charged particle. Possible detection reactions: _ n + n - + p n + p + + n = e, m, t = e, m, t The cross-section for the reaction n e + n e - + p is of the order of cm 2 (per nucleon, E n few MeV). detection efficiency 1 m Iron: 1 km water: det s N a 17 det det 6 10 N s A A d ( Na :area density N A : Avogadro' s number) Neutrino detection requires big and massive detectors (ktons - Mtons) and very high neutrino fluxes (e.g n / yr). In collider experiments fully hermetic detectors allow to detect neutrinos indirectly: - sum up all visible energy and momentum. - attribute missing energy and momentum to neutrino. L3-10

11 Interaction of neutrons and neutrinos Direct neutrino detection Indirect neutrino detection e + e - ( s=181 GeV) W + W - qqmn m 2 hadronic jets + m missing momentum Super-Kamiokande is a 50,000 ton water Cherenkov detector, able observe a couple of thousands of neutrinos per year. e,m n N Cherenkov light L3-11

12 Hadronic cascades Lots of processes involved: Strong, e.m. and weak interactions. Much more complex and larger than electromagnetic cascades. (l I > X 0 ) A hadronic shower has two components: (Grupen) hadronic + electromagnetic charged hadrons p,p,k, nuclear fragmets. breaking up of nuclei (binding energy) neutral pions p 0 2g electromagnetic cascades n p 0 ln E( GeV) 4. 6 neutrons, neutrinos, soft g s, muons example E = 100 GeV: n(p 0 ) 18 invisible energy large fluctuations of visible energy Modest energy resolution of hadronic calorimeters. L3-12

13 Introduction to Scintillators scintillator Light guide (optional) photodetector Energy deposition by an ionizing particle or photon (g) generation transmission detection of scintillation light Two categories Organic scintillators (crystals, plastics or liquid solutions) Inorganic (crystalline structure) Up to photons per MeV Low Z (not good for photoeffect) Low density ~1g/cm 3 Doped, large choice of emission wavelength ns decay times Relatively inexpensive Medium Rad. Hard (10 kgy/year) Used in hadr./e.m. calorimetry, as trigger counters, for lab tests More light (up to ph/mev) High Z, high density Rel. expensive Used in e.m. calorimetry Don t confuse scintillators with lead glass! The light generation in lead glass is actually based on the Cherenkov effect. Lead glass is the poor man s crystal. High density, but little light output. L3-13

14 Organic scintillation mechanism The organic scintillation mechanism is based on the pi-electrons (molecular orbitals) of the benzene ring (C 6 H 6 ). Molecular states (pi orbitals) singlet states ionization energy S s ultra fast S 2 triplet states S 1 nonradiative T 2 fast T 1 phosphorescence >10-4 s slow S 0 L3-14

15 Organic scintillation mechanism crystals (very rarely used in HEP) Organic scintillators liquids exist as (solutions) (rarely used in HEP) e.g. toluene e.g. p-terphenyl solvent + activator plastics (polymerized solutions) (much used in HEP) a) b) e.g. Butyl-PBD e.g. polyvinlyltoluene (a) or polystyrene (b) L3-15

16 Plastic scintillators Often they consist of a solvent + activator and a secondary fluor as wavelength shifter. Solvent DE = de/dx Dx Activator fluorescence light UV (~300 nm) wavelength shifter ( fluor ) Visible ( >~ 400 nm) fast and local energy transfer via non-radiative dipole-dipole interactions (Förster transfer). radiative transfer A fluor has shifted absorption and emission spectra. The difference of the two peaks is called Stokes shift L3-16

17 Two dopant scheme for plastic scintillators Abs. and emission spectra Förster (non-radiative) Radiative, UV Emission, blue L3-17

18 Some examples of commercial plastic scintillators. There are just two main suppliers: (Saint Gobain, FR/US) or ELJEN (US) which deliver rather comparable products. blue UV 8000 photons / MeV L3-18

19 Scintillators readout Readout has to be adapted to geometry, granularity and emission spectrum of scintillator. Geometrical adaptation: Light guides: transfer by total internal reflection (+outer reflector) fish tail adiabatic Wavelength shifter (WLS) bars / fibres small air gap WLS green Photo detector blue (secondary) scintillator UV (primary) primary particle L3-19

20 Most common applications of organic scintillators Large volume liquid or solid detectors neutron detection underground experiments sampling calorimeters (HCAL in CMS or ATLAS, etc.), trigger counters, TOF counters, Fibre tracking (see below) Michigan University: neutron wall. The flat-sided glass tubes contain liquid scintillator. Scintillating tiles of CMS HCAL. Plastic scintillators in various shapes (Saint Gobain) L3-20

21 Scintillating fibres Working principle of scintillating plastic fibres : Double cladding system (developed by CERN RD7) d 0.5 (1 cos ) 5.3% 4p core polystyrene n=1.59 typ. 25 mm scintillating core polystyrene n=1.59 typically <1 mm cladding (PMMA) n= mm cladding (PMMA) n=1.49 fluorinated outer cladding n= mm light transport by total internal reflection n2 arccos n 1 n 1 Trapping fraction n 2 d 0.5 (1 cos ) 3% 4p There are also square fibres. Their trapping fraction is slightly higher (additional angular phase space), but corners are problematic for light transport. (per side) L3-21

22 luminescence quenching excitation (by charged particle) Inorganic scintillation mechanism Based on band structure of a crystal. Does not work for liquids or gases. conduction band exciton band electron activation centres (impurities) scintillation ( nm, E g < E g ) hole valence band traps E g Band gap E g should be large (>3 ev) to ensure that crystal is transparent. Warning: sometimes 2 time constants: fast recombination (ns-ms) from activation centers delayed recombination due to trapping (ms-ms) L3-22

23 Inorganic scintillators (Grupen) ~75000 PbWO 4 crystals in the CMS electromagnetic calorimeter L3-23

24 Please wake up New topic: Photodetection (Detection of light in the optical domain) L3-24

25 Photodetection in HEP The classical domains of application Calorimetry Readout of organic and inorganic scintillators, lead glass, scint. or quartz fibres Blue/VIS, usually 10s 10000s of photons Particle Identification Detection of Cherenkov light UV/blue, single photons Time Of Flight Usually readout of organic scintillators (not competitive at high momenta) or Cherenkov radiators Tracking Readout of scintillating fibres blue/vis, few photons L3-25

26 Basics of photon detection Purpose: Convert light into detectable electronic signal (we are not covering photographic emulsions!) Principle: Use photoelectric effect to convert photons (g) to photoelectrons (pe) Details depend on the type of the photosensitive material (see below). Photon detection involves often materials like K, Na, Rb, Cs (alkali metals). They have the smallest electronegativity highest tendency to release electrons. L3-26

27 Basics of photodetection Many photosensitive materials are semiconductors, but photoeffect can also be observed from gases and liquids. P [W] semiconductor vacuum e - E A = electron affinity photodet I [A] E g h (Photonis) N γ Internal photoeffect: E g > E g photodet External photoeffect: E g > E g + E A N pe L3-27

28 Basics of photon detection Requirements on photodetectors High sensitivity, usually expressed as: quantum efficiency or radiant sensitivity S (ma/w), with QE % QE % S ma/ W 124 l( nm) N pe N g QE can be >100% (for high energetic photons)! Good Linearity: Output signal light intensity, over a large dynamic range (critical e.g. in calorimetry (energy measurement). Fast Time response: Signal is produced instantaneously (within ns), low jitter (<ns), no afterpulses Low intrinsic noise. A noise-free detector doesn t exist. Thermally created photoelectrons represent the lower limit for the noise rate A o T 2 exp(-ew ph /kt). In many detector types, noise is dominated by other sources. + many more (size, fill factor, radiation hardness, cost, tolerance/immunity to B-fields ) L3-28

29 Frequently used photosensitive materials / photocathodes TEA quartz borosilicate glass normal window glass TMAE, CsI begin of arrow indicates threshold Ultra Violet (UV) Visible Bialkali K 2 CsSb Infra Red (IR) Multialkali NaKCsSb GaAs E [ev] Si (1100 nm) l [nm] NaF, MgF 2, LiF, CaF 2 Cut-off limits of window materials Remember : E[eV] 1239/l[nm] Almost all photosensitive materials are very reactive (alkali metals). Operation only in vacuum or extremely clean gas. Exception: Silicon, CsI. L3-29

30 Quantum Efficiency [%] Latest generation of high performance photocathodes QE Comparison of semitransparent bialkali QE UBA:43% x1.6 Ultra Bialkali available only for small metal channel dynode tubes Example Data for UBA : R SBA : R STD : R SBA:35% STD:26% x1.3 Super Bialkali available for a couple of standard tubes up to Wavelength [nm] L3-30

31 Photo-multiplier tubes (PMT s) Basic principle: Photo-emission from photo-cathode Secondary emission from N dynodes: - dynode gain g 3-50 (function of incoming electron energy E); - total gain M: N M g i i 1 Example: 10 dynodes with g = 4 M = (Hamamatsu) Very sensitive to magnetic fields, even to earth magnetic field (30-60 mt = Gauss). Shielding required (mu-metal). pe ( L3-31

32 SE coefficient d Counts Counts SE coefficient d Gain fluctuations of PMT s Mainly determined by the fluctuations of the number of secondary electrons m i emitted from the dynodes; Poisson distribution: P( n, m i ) n mi e n! m i GaP(Cs) dynodes E A <0 Standard deviation: s n m i m m i i 1 m i (Photonis) fluctuations dominated by 1 st dynode gain m 1 = d 1 CuBe dynodes E A >0 e energy 1 pe 1 pe 2 pe (Photonis) Pedestal noise (Photonis) (H. Houtermanns, NIM 112 (1973) 121) 3 pe Electron energy Pulse height Pulse height L3-32

33 Multi-anode and flat-panel PMT s (Hamamatsu) Cherenkov rings from 3 GeV/c p through aerogel Multi-anode PMT (Hamamatsu) Up to 8 8 channels (2 2 mm 2 each); Size: mm 2 ; Bialkali PC: QE l max = 400 nm; Gain ; Gain uniformity typ. 1 : 2.5; Cross-talk typ. 2% Flat-panel (Hamamatsu H8500): 8 x 8 channels (5.8 x 5.8 mm2 each) Excellent surface coverage (89%) 50 mm (T. Matsumoto et al., NIMA 521 (2004) 367) (Hamamatsu) L3-33

34 50 mm Available with up to 1024 (32 x 32) channels (1.6 x 1.6 mm 2 ) Micro Channel Plate (MCP) based PMTs Window/Faceplate Photocathode Dual MCP MCP-OUT Pulse Anode photon Photoelectron Gain ~ 10 6 DV ~ 200V DV~ 2000V DV ~ 200V MCPs are usually based on glass disks, with lots of aligned pores. The surface of the pores are metal coated. Gain stage and detection are decoupled lots of potential and freedom for MA-PMTs: Anode can be easily segmented in application specific way. Typical secondary yield is 2 For 40:1 L:D there are typically 10 strikes (2 10 ~ 10 3 gain per single plate) Pore sizes range from <10 to 25 mm. Small distances small TTS and good immunity to B-field Dual MCP Anode & Pins Ceramic Insulators L3-34

35 Light absorption in Silicon At large l, temperature effects become important ( deg/ccd.html L3-35

36 Solid-state photon detectors (Si) Photodiodes (PIN diode) P(I)N type p layer very thin (<1 mm), as visible light is rapidly absorbed by silicon High QE l 700nm) Gain = 1 Avalanche photodiode (APD) High reverse bias voltage: typ. few 100 V Special doping profile high internal field (>10 5 V/cm) e and h avalanche multiplication Avalanche must stop due to statistical fluctuations. Gain: typ. O(100) g p + i(n) n + h e Rel. high gain fluctuations (excess noise from the avalanche). CMS ECAL APD: ENF = Very high sensitivity on temp. and bias voltage DG = 3.1%/V and -2.4 %/K Hamamatsu S8148. ( pieces used in CMS barrel ECAL). L3-36

37 J. Haba, RICH2007 J. Haba, RICH2007 J. Haba, RICH2007 Solid-state Geiger mode Avalanche Photodiode (G-APD) How to obtain higher gain (= single photon detection) without suffering from excessive noise? Operate APD cell in Geiger mode (= full discharge), however with (passive) quenching. Photon conversion + avalanche short-circuits the diode. L3-37

38 J. Haba, RICH2007 J. Haba, RICH2007 Solid-state Geiger mode Avalanche Photodiode (G-APD) t = R S C D (sub ns) I D I max ~(V BIAS -V BD )/R Q t = R Q C D 10s of ns Gain = Q / e = I max t / e = (V BIAS -V BD )C D / e = DV C D / e G ~ at reasonable bias voltage (<100 V) Sample of 3 G-APDs Sample of 3 G-APDs L3-38

39 Multi pixel G-APD, called G-APD, MPPC, SiPM, 1mm 100 several 1000 pix / mm 2 GM-APD Bias bus Quench resistor Sizes up to 6 6 mm 2 now standard. Quench resistor g V bias 20 x 20 pix g 1 g 2 g 3 g GM-APD Q Q 2Q Quasi-analog detector allows photon counting with a clearly quantized signal L3-39

40 Hamamatsu catalog Multi pixel G-APD = G-APD, MPPC, SiPM, You cannot get "something for nothing G-APD show dark noise rate in the O(100 khz MHz / mm 2 ) range. The gain is temperature dependent O(<5% / K) The signal linearity is limited The price is (still too) high N fired N 0 x 1 exp N 0 L3-40

41 SiPM designs (just examples) Hamamatsu HPK ( 25x25µm 2, 50x50µm 2, 100x100µm 2 pixel size 1x1mm 2 FBK-IRST 50x50µm 2 pixel size 1x1mm 2 2x2mm 2 3x3mm 2 4x4mm 2 3x3mm 2 Arrays 1x4mm 2 1x4 channels 4x4mm 2 6x6 mm 2 2x2 channels 2x2 channels SensL ( 20x20µm 2, 35x35µm 2, 50x50µm 2, 100x100µm 2 pixel size 3x3 cm 2 8x8 channels 3.16x3.16mm 2 4x4 channels 3.16x3.16mm 2 4x4 channels 6 x 6 cm 2 16x16 channels L3-41

42 Gaseous photodetectors: A few implementations... Proven technology: Cherenkov detectors in ALICE, HADES, COMPASS, J-LAB. Many m 2 of CsI photocathodes Micro Pattern Structures (GEM) + CsI HBD (RICH) of PHENIX. R&D: Thick GEM structures Visible PC (bialkali) Sealed gaseous devices HV photocathode CsI on readout pads CsI on multi-gem structure Sealed gaseous photodetector with bialkali PC. (Weizmann Inst., Israel) L3-42

43 BACK-UP SLIDES L3-43

44 Dark counts due to Thermal/tunneling : thermal/ tunneling carrier generation in the bulk or in the surface depleted region around the junction After-pulses : carriers trapped during the avalanche discharging and then released triggering a new avalanche during a period of several 100 ns after the breakdown Optical cross-talk: 10 5 carriers in an avalanche plasma emit on average 3 photons with an energy higher than 1.14 ev (A. Lacaita et al. IEEE TED 1993). These photons can trigger an avalanche in an adjacent µcell. Limit gain, increase threshold add trenches btw mcells 1.E+07 th=0.5pe dark count rate (Hz) 1.E+06 1.E+05 1.E+04 1.E+03 N. Dinu & al, al, NIM A A 572 (2007) FBK-irst device 2007 production DV=0.5VDV=1V DV=1.5V DV=2V DV=2.5 V DV=3 V 1.E threshold (mv) L3-44

45 Digital SiPM Implementation of SiPMs in a CMOS process allows adding lots of functionality 1 digital cell (area depends on type, e.g. 30 x 50 mm) Philips v Different from analog SiPMs: Upon the detection of a photon, the avalanche is actively quenched using a dedicated transistor, and a different transistor is used to quickly recharge the diode back to its sensitive state. L3-45

46 Digital SiPM Compared to the analog technology, the digital one (offered by Philips) has a number of advantages + Integration of bias supply, amp, TDC, counter + Fast active quenching no afterpulses + Possibility to de-activate noisy cells potentially lower dark noise + Reduced sensitivity to voltage and temperature variations + Compactness + Possibility to add local intelligence 38.5 x 38.5 mm 2 Front and back sides of a 64 channel digital tile (DPC ) or (DPC ) problems shared with analog o High dark noise (a discharging cell doesn t know whether it is digital or analog) o Signal saturation (limited number of cells) Philips and also has some drawbacks The local electronics is a source of heat cooling advisable The readout functionality is designed into the sensor. In case of mismatch with the needs, relatively expensive modifications of the sensor/fpga may be required. L3-46

47 Hybrid Photon Detectors (HPD s) Basic principle: photocathode Combination of vacuum photon detectors and solidstate technology; Optical window, (semitransparent) photo-cathode; Electron optics (optional: demagnification) Charge Gain: achieved in one step by energy dissipation of kev pe s in solid-state detector anode; this results in low gain fluctuations; focusing electrodes electron DV Encapsulation of Si-sensor in the tube implies: o compatibility with high vacuum technology (low outgassing, high T bake-out cycles); silicon sensor Energy loss ev th in (thin) ohmic contact n + p + n W Si e-h = 3.6 ev o o internal (for speed and fine segmentation) or external connectivity to read-out electronics; heat dissipation issues; M s M e( DV Vth ) W Si F M DV = 20 kv M ~ 5000 F = Fano factor F Si ~ 0.1 L3-47

48 Pedestal cut Hybrid Photon Detectors (HPD s) 10-inches (25.4 cm) 1 p.e. 2 p.e. pulse height signals of 1 Si pad HV HPD = 26 kv 3 p.e. 4 p.e. 5 p.e. 10-inch prototype HPD (CERN) for Air Shower Telescope CLUE. pulse height (ADC counts) Photon counting. Continuum due to electron back scattering. L3-48

49 72mm active Pixel-HPD s for LHCb RICH detectors 50mm Cross-focused electron optics pixel array sensor bump-bonded to binary electronic chip, developed at CERN 8192 pixels of mm. specially developed high T bump-bonding; Flip-chip assembly, tube encapsulation (multialkali PC) performed in industry (VTT, Photonis/DEP) T. Gys, NIM A 567 (2006) Pixel-HPD anode During commissioning: illumination of 144 tubes by beamer. In total : 484 tubes. L3-49

50 Gaseous Photodetectors Principle: (A) Ionize photosensitive molecules, admixed to the counter gas (TMAE, TEA); e.g. CH 4 + TEA or (B) release photoelectron from a solid photocathode (CsI, bialkali...); Then use free photoelectron to trigger a Townsend avalanche Gain TEA, TMAE, CsI work only in deep UV region. Bialkali works in visible domain, however requires VERY clean gases. Long term operation in a real detector not yet demonstrated. Thin CsI coating on cathode pads Usual issues: How to achieve high gain (10 5 )? How to control ion feedback and light emisson from avalanche? How to purify gas and keep it clean? How to control aging? L3-50

51 Side remark Bremsstrahlung plays an important role in accelerators, but essentially only in circular e ± machines. Here it is called Synchrotron radiation e ± F R Magnetic field forces particles on a circle permanent acceleration towards the centre of the circle. Radiated power: P r P P r, e r, p e 2 c 1 E p 0 m R m m 4 e 4 p Negative aspect: The radiated energy per turn can eat up the gained energy by acceleration and so limit the achievable energy. Famous example: LEP (e + e - collider). 2pR DEturn Pr dt Pr c E LEP ~100 GeV. R LEP ~4300 m DE turn ~ 2 GeV Positive aspect: The radiated energy is extremely forward peaked (Lorentz transformed) and can be used as very bright and intense photon source, e.g. for material studies. See e.g. ESRF ( Dipole radiation Forward peaked SR L3-51

52 A few more words on muons (at very high energies) Muons interact electromagnetically (like the e +,e -, but due to its high mass, direct Bremsstrahlung (~E/m 2 ) is strongly suppressed. There are other radiative processes Photo-nuclear interactions Z,A Pair production Z,A Like for Bremsstrahlung, they scale with E de b( Z, A, E m ) E m dx L3-52

53 Total energy loss of muons (in different materials) de dx a b E m ionization radiative L3-53

54 substrate Basics of photon detection Opaque photocathode g Light absorption in photocathode 0.4 l A = 1/ e - PC Red light (l 600 nm) cm -1 l A 60 nm Blue light (l 400 nm) cm -1 l A 25 nm Semitransparent photocathode g Detector window Blue light is stronger absorped than red light! PC e - Make semitransparent photocathode just as thick as necessary! L3-54