[{"data":1,"prerenderedAt":869},["ShallowReactive",2],{"content-query-CoXOKrZuLp":3},{"_path":4,"_dir":5,"_draft":6,"_partial":6,"_locale":7,"title":8,"description":9,"heading":10,"prompt":11,"tags":15,"files":18,"nav":6,"gallery":19,"body":21,"_type":862,"_id":863,"_source":864,"_file":865,"_stem":866,"_extension":867,"sitemap":868},"/tools/lorentzian-peak-fit","tools",false,"","Lorentzian Peak Fit Calculator","Fit Lorentzian peaks online from Excel or CSV data. Estimate center, width, amplitude, and residual diagnostics with AI.","Lorentzian Peak Fit",{"prefix":12,"label":13,"placeholder":14},"Fit Lorentzian peaks","Describe the peak fitting analysis you want to run","e.g. fit 4 Lorentzian peaks to NMR spectrum, report chemical shifts, linewidths, and relative intensities",[16,17],"statistics","science",true,[20],"/img/tools/lorentzian-peak-fit.png",{"type":22,"children":23,"toc":851},"root",[24,33,102,142,148,220,226,343,348,354,504,510,634,640,708,714,748,754,771,794,818,841],{"type":25,"tag":26,"props":27,"children":29},"element","h2",{"id":28},"what-is-a-lorentzian-peak",[30],{"type":31,"value":32},"text","What Is a Lorentzian Peak?",{"type":25,"tag":34,"props":35,"children":36},"p",{},[37,39,45,47,52,54,65,67,72,74,79,81,86,88,93,95,100],{"type":31,"value":38},"A ",{"type":25,"tag":40,"props":41,"children":42},"strong",{},[43],{"type":31,"value":44},"Lorentzian",{"type":31,"value":46}," (also called a ",{"type":25,"tag":40,"props":48,"children":49},{},[50],{"type":31,"value":51},"Cauchy",{"type":31,"value":53}," distribution in statistics) is a symmetric peak-shaped function defined as ",{"type":25,"tag":40,"props":55,"children":56},{},[57,59],{"type":31,"value":58},"L(x) = A × γ² / ",{"type":25,"tag":60,"props":61,"children":62},"span",{},[63],{"type":31,"value":64},"(x − x₀)² + γ²",{"type":31,"value":66},", where ",{"type":25,"tag":40,"props":68,"children":69},{},[70],{"type":31,"value":71},"A",{"type":31,"value":73}," is the peak amplitude, ",{"type":25,"tag":40,"props":75,"children":76},{},[77],{"type":31,"value":78},"x₀",{"type":31,"value":80}," is the peak center, and ",{"type":25,"tag":40,"props":82,"children":83},{},[84],{"type":31,"value":85},"γ",{"type":31,"value":87}," is the half-width at half maximum (HWHM). The ",{"type":25,"tag":40,"props":89,"children":90},{},[91],{"type":31,"value":92},"full width at half maximum is FWHM = 2γ",{"type":31,"value":94},". Unlike the Gaussian, which falls off as a squared exponential and has negligible tails beyond a few widths, the Lorentzian falls off as 1/x² and has ",{"type":25,"tag":40,"props":96,"children":97},{},[98],{"type":31,"value":99},"heavy algebraic tails",{"type":31,"value":101}," — a defining feature that makes it physically correct for many spectroscopic systems.",{"type":25,"tag":34,"props":103,"children":104},{},[105,107,112,114,119,121,126,128,133,135,140],{"type":31,"value":106},"Lorentzian lineshapes arise from ",{"type":25,"tag":40,"props":108,"children":109},{},[110],{"type":31,"value":111},"homogeneous broadening",{"type":31,"value":113}," mechanisms: natural lifetime broadening (governed by the Heisenberg uncertainty principle), collisional broadening in gases, and spin relaxation in NMR. In ",{"type":25,"tag":40,"props":115,"children":116},{},[117],{"type":31,"value":118},"NMR spectroscopy",{"type":31,"value":120},", every peak in a ¹H, ¹³C, or ³¹P spectrum is a Lorentzian — the linewidth (in Hz) is directly related to the transverse relaxation time T₂ by FWHM = 1/(πT₂). In ",{"type":25,"tag":40,"props":122,"children":123},{},[124],{"type":31,"value":125},"Raman spectroscopy",{"type":31,"value":127},", phonon modes in crystalline materials produce Lorentzian peaks whose width reports on phonon lifetime and defect density. In ",{"type":25,"tag":40,"props":129,"children":130},{},[131],{"type":31,"value":132},"resonance physics",{"type":31,"value":134},", the cross-section near a resonance (e.g. optical cavity resonance, atomic absorption line) follows a Lorentzian described by the ",{"type":25,"tag":40,"props":136,"children":137},{},[138],{"type":31,"value":139},"Breit-Wigner formula",{"type":31,"value":141},". In these contexts, fitting a Gaussian instead of a Lorentzian systematically underestimates the tails and over-estimates narrow linewidths.",{"type":25,"tag":26,"props":143,"children":145},{"id":144},"how-it-works",[146],{"type":31,"value":147},"How It Works",{"type":25,"tag":149,"props":150,"children":151},"ol",{},[152,177,193],{"type":25,"tag":153,"props":154,"children":155},"li",{},[156,161,163,168,170,175],{"type":25,"tag":40,"props":157,"children":158},{},[159],{"type":31,"value":160},"Upload your data",{"type":31,"value":162}," — provide a CSV or Excel file with an ",{"type":25,"tag":40,"props":164,"children":165},{},[166],{"type":31,"value":167},"x",{"type":31,"value":169}," column (chemical shift in ppm, wavenumber in cm⁻¹, frequency in Hz, or any continuous variable) and a ",{"type":25,"tag":40,"props":171,"children":172},{},[173],{"type":31,"value":174},"y",{"type":31,"value":176}," column (intensity, counts, absorbance). One row per data point.",{"type":25,"tag":153,"props":178,"children":179},{},[180,185,187],{"type":25,"tag":40,"props":181,"children":182},{},[183],{"type":31,"value":184},"Describe the analysis",{"type":31,"value":186}," — e.g. ",{"type":25,"tag":188,"props":189,"children":190},"em",{},[191],{"type":31,"value":192},"\"fit 4 Lorentzian peaks to the NMR spectrum; x column is 'ppm', y is 'intensity'; report chemical shifts, FWHM linewidths, and relative integrals; plot individual components and sum\"",{"type":25,"tag":153,"props":194,"children":195},{},[196,201,203,210,212,218],{"type":25,"tag":40,"props":197,"children":198},{},[199],{"type":31,"value":200},"Get full results",{"type":31,"value":202}," — the AI writes Python code using ",{"type":25,"tag":204,"props":205,"children":207},"a",{"href":206},"https://docs.scipy.org/doc/scipy/reference/generated/scipy.optimize.curve_fit.html",[208],{"type":31,"value":209},"scipy.optimize.curve_fit",{"type":31,"value":211}," to fit the multi-Lorentzian model and ",{"type":25,"tag":204,"props":213,"children":215},{"href":214},"https://plotly.com/python/",[216],{"type":31,"value":217},"Plotly",{"type":31,"value":219}," to render the deconvoluted spectrum with filled components, total fit overlay, and parameter table",{"type":25,"tag":26,"props":221,"children":223},{"id":222},"required-data-format",[224],{"type":31,"value":225},"Required Data Format",{"type":25,"tag":227,"props":228,"children":229},"table",{},[230,254],{"type":25,"tag":231,"props":232,"children":233},"thead",{},[234],{"type":25,"tag":235,"props":236,"children":237},"tr",{},[238,244,249],{"type":25,"tag":239,"props":240,"children":241},"th",{},[242],{"type":31,"value":243},"Column",{"type":25,"tag":239,"props":245,"children":246},{},[247],{"type":31,"value":248},"Description",{"type":25,"tag":239,"props":250,"children":251},{},[252],{"type":31,"value":253},"Example",{"type":25,"tag":255,"props":256,"children":257},"tbody",{},[258,302],{"type":25,"tag":235,"props":259,"children":260},{},[261,271,276],{"type":25,"tag":262,"props":263,"children":264},"td",{},[265],{"type":25,"tag":266,"props":267,"children":269},"code",{"className":268},[],[270],{"type":31,"value":167},{"type":25,"tag":262,"props":272,"children":273},{},[274],{"type":31,"value":275},"Independent variable",{"type":25,"tag":262,"props":277,"children":278},{},[279,285,287,293,294,300],{"type":25,"tag":266,"props":280,"children":282},{"className":281},[],[283],{"type":31,"value":284},"7.25",{"type":31,"value":286},", ",{"type":25,"tag":266,"props":288,"children":290},{"className":289},[],[291],{"type":31,"value":292},"7.26",{"type":31,"value":286},{"type":25,"tag":266,"props":295,"children":297},{"className":296},[],[298],{"type":31,"value":299},"7.27",{"type":31,"value":301}," … (ppm)",{"type":25,"tag":235,"props":303,"children":304},{},[305,313,318],{"type":25,"tag":262,"props":306,"children":307},{},[308],{"type":25,"tag":266,"props":309,"children":311},{"className":310},[],[312],{"type":31,"value":174},{"type":25,"tag":262,"props":314,"children":315},{},[316],{"type":31,"value":317},"Signal intensity",{"type":25,"tag":262,"props":319,"children":320},{},[321,327,328,334,335,341],{"type":25,"tag":266,"props":322,"children":324},{"className":323},[],[325],{"type":31,"value":326},"0.02",{"type":31,"value":286},{"type":25,"tag":266,"props":329,"children":331},{"className":330},[],[332],{"type":31,"value":333},"0.45",{"type":31,"value":286},{"type":25,"tag":266,"props":336,"children":338},{"className":337},[],[339],{"type":31,"value":340},"1.00",{"type":31,"value":342}," (a.u.)",{"type":25,"tag":34,"props":344,"children":345},{},[346],{"type":31,"value":347},"Any column names work — describe them in your prompt. NMR data is typically exported from spectrometer software as a two-column CSV. For Raman or FTIR, export the spectral window of interest.",{"type":25,"tag":26,"props":349,"children":351},{"id":350},"interpreting-the-results",[352],{"type":31,"value":353},"Interpreting the Results",{"type":25,"tag":227,"props":355,"children":356},{},[357,373],{"type":25,"tag":231,"props":358,"children":359},{},[360],{"type":25,"tag":235,"props":361,"children":362},{},[363,368],{"type":25,"tag":239,"props":364,"children":365},{},[366],{"type":31,"value":367},"Parameter",{"type":25,"tag":239,"props":369,"children":370},{},[371],{"type":31,"value":372},"What it means",{"type":25,"tag":255,"props":374,"children":375},{},[376,392,408,424,440,456,472,488],{"type":25,"tag":235,"props":377,"children":378},{},[379,387],{"type":25,"tag":262,"props":380,"children":381},{},[382],{"type":25,"tag":40,"props":383,"children":384},{},[385],{"type":31,"value":386},"Peak center x₀",{"type":25,"tag":262,"props":388,"children":389},{},[390],{"type":31,"value":391},"Position of maximum — chemical shift (ppm), wavenumber (cm⁻¹), frequency (Hz)",{"type":25,"tag":235,"props":393,"children":394},{},[395,403],{"type":25,"tag":262,"props":396,"children":397},{},[398],{"type":25,"tag":40,"props":399,"children":400},{},[401],{"type":31,"value":402},"Amplitude A",{"type":25,"tag":262,"props":404,"children":405},{},[406],{"type":31,"value":407},"Peak height at x₀",{"type":25,"tag":235,"props":409,"children":410},{},[411,419],{"type":25,"tag":262,"props":412,"children":413},{},[414],{"type":25,"tag":40,"props":415,"children":416},{},[417],{"type":31,"value":418},"γ (HWHM)",{"type":25,"tag":262,"props":420,"children":421},{},[422],{"type":31,"value":423},"Half-width at half maximum — FWHM = 2γ",{"type":25,"tag":235,"props":425,"children":426},{},[427,435],{"type":25,"tag":262,"props":428,"children":429},{},[430],{"type":25,"tag":40,"props":431,"children":432},{},[433],{"type":31,"value":434},"FWHM",{"type":25,"tag":262,"props":436,"children":437},{},[438],{"type":31,"value":439},"Full linewidth — inversely related to T₂ relaxation time in NMR",{"type":25,"tag":235,"props":441,"children":442},{},[443,451],{"type":25,"tag":262,"props":444,"children":445},{},[446],{"type":25,"tag":40,"props":447,"children":448},{},[449],{"type":31,"value":450},"Integrated area = Aπγ",{"type":25,"tag":262,"props":452,"children":453},{},[454],{"type":31,"value":455},"Proportional to the number of nuclei (NMR) or scatterers (Raman)",{"type":25,"tag":235,"props":457,"children":458},{},[459,467],{"type":25,"tag":262,"props":460,"children":461},{},[462],{"type":25,"tag":40,"props":463,"children":464},{},[465],{"type":31,"value":466},"Relative area (%)",{"type":25,"tag":262,"props":468,"children":469},{},[470],{"type":31,"value":471},"Ratio of integrals — corresponds to proton count ratio in ¹H NMR",{"type":25,"tag":235,"props":473,"children":474},{},[475,483],{"type":25,"tag":262,"props":476,"children":477},{},[478],{"type":25,"tag":40,"props":479,"children":480},{},[481],{"type":31,"value":482},"Q-factor = x₀ / FWHM",{"type":25,"tag":262,"props":484,"children":485},{},[486],{"type":31,"value":487},"Resonance quality factor for cavity/resonator measurements",{"type":25,"tag":235,"props":489,"children":490},{},[491,499],{"type":25,"tag":262,"props":492,"children":493},{},[494],{"type":25,"tag":40,"props":495,"children":496},{},[497],{"type":31,"value":498},"Residuals",{"type":25,"tag":262,"props":500,"children":501},{},[502],{"type":31,"value":503},"Flat random scatter = good fit; systematic structure = missing peak or wrong model",{"type":25,"tag":26,"props":505,"children":507},{"id":506},"example-prompts",[508],{"type":31,"value":509},"Example Prompts",{"type":25,"tag":227,"props":511,"children":512},{},[513,529],{"type":25,"tag":231,"props":514,"children":515},{},[516],{"type":25,"tag":235,"props":517,"children":518},{},[519,524],{"type":25,"tag":239,"props":520,"children":521},{},[522],{"type":31,"value":523},"Scenario",{"type":25,"tag":239,"props":525,"children":526},{},[527],{"type":31,"value":528},"What to type",{"type":25,"tag":255,"props":530,"children":531},{},[532,549,566,583,600,617],{"type":25,"tag":235,"props":533,"children":534},{},[535,540],{"type":25,"tag":262,"props":536,"children":537},{},[538],{"type":31,"value":539},"NMR deconvolution",{"type":25,"tag":262,"props":541,"children":542},{},[543],{"type":25,"tag":266,"props":544,"children":546},{"className":545},[],[547],{"type":31,"value":548},"fit Lorentzian peaks to ¹H NMR; x is 'ppm', y is 'intensity'; report chemical shifts and relative integrals",{"type":25,"tag":235,"props":550,"children":551},{},[552,557],{"type":25,"tag":262,"props":553,"children":554},{},[555],{"type":31,"value":556},"Raman band analysis",{"type":25,"tag":262,"props":558,"children":559},{},[560],{"type":25,"tag":266,"props":561,"children":563},{"className":562},[],[564],{"type":31,"value":565},"fit 3 Lorentzian peaks to Raman spectrum 1200–1700 cm⁻¹; report band positions, FWHM, and areas",{"type":25,"tag":235,"props":567,"children":568},{},[569,574],{"type":25,"tag":262,"props":570,"children":571},{},[572],{"type":31,"value":573},"Linewidth / T₂",{"type":25,"tag":262,"props":575,"children":576},{},[577],{"type":25,"tag":266,"props":578,"children":580},{"className":579},[],[581],{"type":31,"value":582},"fit Lorentzian to NMR peak; report FWHM in Hz; compute T₂ = 1/(π × FWHM)",{"type":25,"tag":235,"props":584,"children":585},{},[586,591],{"type":25,"tag":262,"props":587,"children":588},{},[589],{"type":31,"value":590},"Overlapping NMR peaks",{"type":25,"tag":262,"props":592,"children":593},{},[594],{"type":25,"tag":266,"props":595,"children":597},{"className":596},[],[598],{"type":31,"value":599},"deconvolute overlapping doublet at 3.5 ppm; fit 2 Lorentzians with equal widths; coupling constant from peak separation",{"type":25,"tag":235,"props":601,"children":602},{},[603,608],{"type":25,"tag":262,"props":604,"children":605},{},[606],{"type":31,"value":607},"Voigt fit comparison",{"type":25,"tag":262,"props":609,"children":610},{},[611],{"type":25,"tag":266,"props":612,"children":614},{"className":613},[],[615],{"type":31,"value":616},"fit Lorentzian, Gaussian, and Voigt to same peak; compare AIC and residuals; report best-fit FWHM",{"type":25,"tag":235,"props":618,"children":619},{},[620,625],{"type":25,"tag":262,"props":621,"children":622},{},[623],{"type":31,"value":624},"Baseline + peaks",{"type":25,"tag":262,"props":626,"children":627},{},[628],{"type":25,"tag":266,"props":629,"children":631},{"className":630},[],[632],{"type":31,"value":633},"fit 4 Lorentzian peaks plus cubic spline baseline to FTIR spectrum; report integrated absorbances",{"type":25,"tag":26,"props":635,"children":637},{"id":636},"assumptions-to-check",[638],{"type":31,"value":639},"Assumptions to Check",{"type":25,"tag":641,"props":642,"children":643},"ul",{},[644,661,671,688,698],{"type":25,"tag":153,"props":645,"children":646},{},[647,652,654,659],{"type":25,"tag":40,"props":648,"children":649},{},[650],{"type":31,"value":651},"Lorentzian is the correct model",{"type":31,"value":653}," — Lorentzian is appropriate for homogeneously broadened lines (NMR T₂ relaxation, natural linewidth, phonon lifetime). For inhomogeneously broadened lines (e.g. solid-state NMR powder patterns, optical absorption bands in solution), a Gaussian or ",{"type":25,"tag":40,"props":655,"children":656},{},[657],{"type":31,"value":658},"Voigt",{"type":31,"value":660}," (Lorentzian ⊗ Gaussian) is more appropriate",{"type":25,"tag":153,"props":662,"children":663},{},[664,669],{"type":25,"tag":40,"props":665,"children":666},{},[667],{"type":31,"value":668},"Baseline subtracted",{"type":31,"value":670}," — Lorentzian tails extend far from the peak center; a sloping or curved baseline significantly distorts the fitted γ and amplitude. Include a polynomial baseline term or pre-process the data",{"type":25,"tag":153,"props":672,"children":673},{},[674,679,681,686],{"type":25,"tag":40,"props":675,"children":676},{},[677],{"type":31,"value":678},"Good initial guesses",{"type":31,"value":680}," — provide approximate peak positions and widths in your prompt: ",{"type":25,"tag":188,"props":682,"children":683},{},[684],{"type":31,"value":685},"\"peaks near 1350 and 1580 cm⁻¹, widths around 20 cm⁻¹\"",{"type":31,"value":687},". Lorentzian fitting is more sensitive to initial guesses than Gaussian fitting due to the heavy tails",{"type":25,"tag":153,"props":689,"children":690},{},[691,696],{"type":25,"tag":40,"props":692,"children":693},{},[694],{"type":31,"value":695},"Avoid fitting noise",{"type":31,"value":697}," — very narrow peaks close to the digital resolution of your spectrometer cannot be reliably fit; the Lorentzian γ cannot be smaller than half the point spacing",{"type":25,"tag":153,"props":699,"children":700},{},[701,706],{"type":25,"tag":40,"props":702,"children":703},{},[704],{"type":31,"value":705},"Far wings contain information",{"type":31,"value":707}," — because Lorentzian tails are heavy, data far from the peak center still influences the fit; ensure the baseline region is correctly modeled",{"type":25,"tag":26,"props":709,"children":711},{"id":710},"related-tools",[712],{"type":31,"value":713},"Related Tools",{"type":25,"tag":34,"props":715,"children":716},{},[717,719,725,727,731,733,738,740,746],{"type":31,"value":718},"Use the ",{"type":25,"tag":204,"props":720,"children":722},{"href":721},"/tools/gaussian-peak-fit",[723],{"type":31,"value":724},"Gaussian Peak Fit",{"type":31,"value":726}," when your peaks are inhomogeneously broadened — optical absorption in solution, chromatography peaks, or any system where a distribution of environments broadens the line. Use the ",{"type":25,"tag":204,"props":728,"children":729},{"href":721},[730],{"type":31,"value":724},{"type":31,"value":732}," and ask for a ",{"type":25,"tag":40,"props":734,"children":735},{},[736],{"type":31,"value":737},"Voigt profile",{"type":31,"value":739}," (convolution of Gaussian and Lorentzian) when both broadening mechanisms are present and you want the most accurate lineshape model. Use the ",{"type":25,"tag":204,"props":741,"children":743},{"href":742},"/tools/residual-plot",[744],{"type":31,"value":745},"Residual Plot Generator",{"type":31,"value":747}," to compare fit quality between Lorentzian and Gaussian models systematically.",{"type":25,"tag":26,"props":749,"children":751},{"id":750},"frequently-asked-questions",[752],{"type":31,"value":753},"Frequently Asked Questions",{"type":25,"tag":34,"props":755,"children":756},{},[757,762,764,769],{"type":25,"tag":40,"props":758,"children":759},{},[760],{"type":31,"value":761},"How do I know whether to use Lorentzian or Gaussian for my data?",{"type":31,"value":763},"\nThe key diagnostic is the ",{"type":25,"tag":40,"props":765,"children":766},{},[767],{"type":31,"value":768},"peak tails",{"type":31,"value":770},". Plot your data on a linear scale zoomed out 5–10× the FWHM from the peak center. If the signal decays quickly to baseline (roughly Gaussian), use a Gaussian. If the signal persists far into the wings (slow 1/x² decay), use a Lorentzian. For an objective comparison, fit both models to the same peak, plot the residuals, and compare AIC — the model with lower AIC and flatter residuals is correct. Most NMR peaks are Lorentzian; most optical absorption bands in solution are Gaussian.",{"type":25,"tag":34,"props":772,"children":773},{},[774,779,781,785,787,792],{"type":25,"tag":40,"props":775,"children":776},{},[777],{"type":31,"value":778},"What is the Voigt profile and when should I use it?",{"type":31,"value":780},"\nThe ",{"type":25,"tag":40,"props":782,"children":783},{},[784],{"type":31,"value":737},{"type":31,"value":786}," is the convolution of a Gaussian (width σ) and a Lorentzian (width γ) and represents systems with both homogeneous and inhomogeneous broadening — for example, a gas-phase absorption line that has both Doppler broadening (Gaussian) and collisional broadening (Lorentzian). The Voigt FWHM is approximately 0.5346 × FWHM_L + √(0.2166 × FWHM_L² + FWHM_G²). Ask the AI to ",{"type":25,"tag":188,"props":788,"children":789},{},[790],{"type":31,"value":791},"\"fit a Voigt profile and report the Gaussian and Lorentzian width contributions separately\"",{"type":31,"value":793},".",{"type":25,"tag":34,"props":795,"children":796},{},[797,802,804,809,811,816],{"type":25,"tag":40,"props":798,"children":799},{},[800],{"type":31,"value":801},"My NMR peak has a shoulder — is it one Lorentzian or two?",{"type":31,"value":803},"\nA shoulder usually indicates two overlapping peaks. The reliable test is the ",{"type":25,"tag":40,"props":805,"children":806},{},[807],{"type":31,"value":808},"second derivative",{"type":31,"value":810},": a single Lorentzian has one minimum in its second derivative; two overlapping peaks produce two minima (often with the shoulder appearing as a slight inflection). Ask the AI to ",{"type":25,"tag":188,"props":812,"children":813},{},[814],{"type":31,"value":815},"\"compute the second derivative of the spectrum to count peaks before fitting\"",{"type":31,"value":817},". For overlapping NMR doublets, you can also constrain the fit: fix both peaks to have the same width and set their separation equal to the expected J-coupling constant.",{"type":25,"tag":34,"props":819,"children":820},{},[821,826,828,833,835,840],{"type":25,"tag":40,"props":822,"children":823},{},[824],{"type":31,"value":825},"How do I extract the spin relaxation time T₂ from a Lorentzian linewidth?",{"type":31,"value":827},"\nIn NMR, the transverse relaxation time T₂ is related to the natural linewidth by ",{"type":25,"tag":40,"props":829,"children":830},{},[831],{"type":31,"value":832},"T₂ = 1/(π × FWHM_Hz)",{"type":31,"value":834},", where FWHM must be in Hz (not ppm). Convert ppm to Hz by multiplying by the spectrometer frequency: FWHM_Hz = FWHM_ppm × spectrometer_MHz. Ask the AI to ",{"type":25,"tag":188,"props":836,"children":837},{},[838],{"type":31,"value":839},"\"compute T₂ from the Lorentzian FWHM; spectrometer frequency is 400 MHz; FWHM in ppm\"",{"type":31,"value":793},{"type":25,"tag":34,"props":842,"children":843},{},[844,849],{"type":25,"tag":40,"props":845,"children":846},{},[847],{"type":31,"value":848},"Can I fit peaks with unequal baselines on the two sides?",{"type":31,"value":850},"\nAn asymmetric baseline (one side higher than the other) often indicates either a sloping background or peak overlap with a broad underlying component. The Lorentzian model itself is symmetric. Best practice: (1) fit a local linear baseline using baseline-free spectral windows, (2) subtract it, then (3) fit Lorentzians to the baseline-corrected data. Alternatively, include a linear baseline as two additional free parameters in the fit and let the optimizer determine both baseline and peak parameters simultaneously.",{"title":7,"searchDepth":852,"depth":852,"links":853},2,[854,855,856,857,858,859,860,861],{"id":28,"depth":852,"text":32},{"id":144,"depth":852,"text":147},{"id":222,"depth":852,"text":225},{"id":350,"depth":852,"text":353},{"id":506,"depth":852,"text":509},{"id":636,"depth":852,"text":639},{"id":710,"depth":852,"text":713},{"id":750,"depth":852,"text":753},"markdown","content:tools:045.lorentzian-peak-fit.md","content","tools/045.lorentzian-peak-fit.md","tools/045.lorentzian-peak-fit","md",{"loc":4},1775502471785]