Crystallographic Analysis Platform

ChemFitX provides an online XRD crystallographic analysis workflow for researchers who need to inspect powder diffraction patterns, calculate d-spacing, detect diffraction peaks, explore candidate hkl assignments and refine unit-cell parameters from indexed peak positions. The platform combines pattern preprocessing, interactive visualization, automatic candidate-cell searching, indexed unit-cell refinement, optional reference-CIF matching and theoretical-density calculation in one browser-based workspace.

Upload an XRD pattern, choose the appropriate X-ray wavelength, optionally smooth the data or correct the baseline, detect and review reliable peaks, and then use Auto Index & Refine to search candidate unit cells. The platform ranks candidate solutions, proposes tentative hkl assignments, refines the selected cell from peak positions and reports residuals so that the result can be reviewed rather than accepted as an unexplained automatic answer.

The XRD file and optional reference CIF are read and processed locally in your web browser by this application. They are not uploaded to ChemFitX servers during the analysis.

What the Crystallographic Analysis Platform Can Do

Import powder XRD data from CSV, TSV, TXT, XY, XYE, ASC, text DAT and supported XRDML scans.
Visualize intensity versus 2θ with pan, box zoom, zoom in/out, view undo/redo and home/reset controls.
Calculate d-spacing from peak positions using the selected X-ray wavelength and Bragg relation.
Apply optional smoothing using several methods, including Savitzky-Golay, moving average, Gaussian, Lowess/Loess, Fourier low-pass, wavelet, exponential, median, kernel and Whittaker approaches.
Apply optional baseline correction using ALS, polynomial, ModPoly, IModPoly, rubberband, rolling-ball, spline, wavelet, peak-rejection, linear, straight-line or manual baseline methods.
Automatically detect candidate diffraction peaks using adjustable prominence, separation and maximum-peak settings.
Insert or remove peaks manually when expert review indicates that automatic detection needs correction.
Search candidate unit cells and propose tentative hkl assignments with Auto Index & Refine.
Search all seven crystal systems or restrict the search to a selected system.
Rank automatic indexing candidates using the indexed fraction, residual quality and internal candidate score.
Refine indexed unit-cell parameters from accepted hkl assignments using reciprocal-metric least squares.
Optionally refine a constant 2θ zero shift together with the unit cell.
Compare observed and calculated 2θ values and inspect residuals graphically and in a table.
Generate calculated reflection positions from the refined cell within the current diffraction range.
Load a known reference CIF, reuse its cell metadata and match observed peaks to candidate reflections from that reference cell.
Reuse chemical formula and Z from a CIF when those fields are present, or enter the missing values manually.
Calculate theoretical density from refined cell volume, chemical formula and Z.
Customize and reposition hkl labels on the XRD graph.
Export indexed peak data, a refined-cell CIF and a calculated reflection list.

Supported XRD Data and X-ray Wavelengths

Supported file formats

The platform accepts common text-based powder-diffraction formats including CSV, TSV, TXT, XY, XYE, ASC and text DAT files, together with supported XRDML scans. A valid dataset should contain a 2θ axis and intensity values. The application sorts the data by diffraction angle and ignores unusable non-numeric rows.

X-ray wavelength selection

ChemFitX provides preset wavelengths for commonly used laboratory radiation sources, including Cu Kα1, Cu Kα, Cu Kα2, Co Kα, Fe Kα, Cr Kα and Mo Kα. A custom wavelength can also be entered. Correct wavelength selection is essential because d-spacing and every calculated reflection position depend directly on λ.

Recommended XRD Analysis Workflow

Upload the XRD pattern and verify that the 2θ range and intensity trace look correct.
Select the correct X-ray wavelength before calculating d-spacing or attempting indexing.
Apply smoothing only if needed, using the minimum strength that suppresses noise without moving or merging real peaks.
Apply baseline correction only when justified and inspect the corrected pattern before continuing.
Detect peaks automatically, then review the list and add or remove peaks if necessary.
Run Auto Index & Refine with the crystal system set to Auto unless independent information supports a specific system.
Review the candidate-cell ranking, indexed fraction, RMS Δ2θ and hkl assignments before accepting a solution.
Refine the selected unit cell from the accepted indexed reflections and inspect the residual plot and observed-versus-calculated table.
If a known-phase CIF is available, use the optional reference-CIF workflow to compare experimental peaks with that known cell.
Calculate theoretical density only after a refined cell, chemical formula and valid Z are available.
Export the indexed peak table, refined cell CIF or calculated reflection list for documentation and further analysis.

XRD Pattern Visualization and d-Spacing

The Pattern workspace displays the processed XRD intensity trace as a function of 2θ. Cursor readouts report diffraction angle, intensity and d-spacing, while interactive navigation tools allow detailed inspection of crowded peak regions without changing the underlying data.

For first-order powder diffraction, ChemFitX uses the Bragg relation nλ = 2d sinθ with n = 1, so d = λ/(2 sinθ). The d-spacing value is therefore meaningful only when the diffraction-angle calibration and selected X-ray wavelength are correct.

Smoothing XRD Data

Noise reduction can improve visual inspection and peak detection, but smoothing must be used conservatively. The platform offers multiple smoothing approaches so users can choose a method appropriate to their dataset rather than forcing one universal filter.

Any smoothing operation can change weak features if applied too strongly. Excessive smoothing may shift peak maxima, suppress shoulders or merge nearby diffraction peaks. After smoothing, inspect the processed trace against the original pattern before using the peak list for indexing.

Baseline Correction

Background and baseline contributions can affect peak prominence and automatic peak detection. ChemFitX provides several baseline approaches, including asymmetric least squares and manual anchor-point correction, so users can evaluate the baseline appropriate to their pattern.

Baseline subtraction can produce small negative noise values, and the platform intentionally retains them rather than clipping them to zero. Baseline correction should be judged visually because aggressive subtraction can distort low-intensity diffraction features and influence which peaks are subsequently detected.

Peak Detection and Manual Peak Review

Automatic peak detection identifies local maxima using user-adjustable prominence, minimum peak separation and maximum-peak settings. The resulting peak positions are candidates for crystallographic indexing; they are not automatically guaranteed to represent independent Bragg reflections from a single phase.

Manual peak insertion is available because automatic algorithms can miss weak reflections or select artifacts, shoulders or overlapping maxima. For reliable indexing, review the peak list and use only peaks that are scientifically defensible. Peak positions, rather than approximate peak widths, are used by the crystallographic refinement module.

Automatic hkl Indexing and Candidate-Cell Search

Auto Index & Refine searches candidate unit cells from the detected or manually inserted peak positions without requiring an initial cell. When the crystal system is left on Auto, the platform tests cubic, tetragonal, hexagonal, orthorhombic, trigonal/rhombohedral, monoclinic and triclinic models and ranks the strongest candidate solutions.

The search uses a configurable number of lower-angle candidate peaks, a maximum Miller-index range, a peak-matching tolerance and a maximum cell-edge limit. Candidate solutions are compared using the number of matched peaks and residual quality, and similar duplicate cells are filtered so that the candidate table remains reviewable.

Automatic indexing is intrinsically non-unique for many powder patterns. Different cells can explain similar subsets of peak positions, especially when peaks are missing, overlapping, contaminated by secondary phases or affected by instrumental errors. Candidate ranking should therefore be treated as a search aid rather than automatic proof of a unique crystal structure.

Crystal-System Coverage

Indexed unit-cell refinement supports all seven crystal systems: cubic, tetragonal, orthorhombic, hexagonal, trigonal/rhombohedral, monoclinic and triclinic. The platform uses the appropriate reciprocal-metric relations for each system instead of applying a cubic approximation to non-cubic cells.

Cubic

For cubic cells, a = b = c and α = β = γ = 90°. One independent lattice parameter is refined from the indexed reflections.

Tetragonal and hexagonal

Tetragonal and hexagonal cells use independent a and c parameters with their corresponding metric constraints. Hexagonal calculations use γ = 120°.

Orthorhombic

Orthorhombic refinement uses independent a, b and c parameters with all interaxial angles fixed at 90°.

Trigonal / rhombohedral

The rhombohedral implementation uses primitive rhombohedral axes and recovers the direct a and α values from the fitted reciprocal metric.

Monoclinic and triclinic

Monoclinic refinement uses the unique-b setting implemented by the platform. Triclinic refinement uses the full six independent reciprocal-metric coefficients, so it requires a sufficiently independent set of indexed reflections.

Lattice Centering and Reflection Generation

The advanced starting-cell workflow can apply simple lattice-centering filters for primitive (P), body-centered (I), face-centered (F) and A-, B- or C-centered lattices. These filters help restrict calculated reflections when centering is independently known.

Centering rules are not the same as a complete space-group extinction model. The current module does not automatically apply all screw-axis and glide-plane systematic absences, so calculated reflection lists and hkl suggestions should not be interpreted as a complete space-group solution.

Indexed Unit-Cell Refinement

After tentative hkl assignments are reviewed, the Cell Refinement tab fits the unit cell from the accepted peak positions in reciprocal-metric space. The fitted metric is converted back to direct-cell parameters, and calculated 2θ positions are compared with the observed diffraction angles.

The platform reports refined a, b and c values, α, β and γ where applicable, unit-cell volume, RMS Δ2θ, mean absolute Δ2θ and the optional refined zero shift. It also displays an observed-versus-calculated residual table and residual plot so that systematic misfits can be identified.

Optional 2θ zero-shift refinement

A constant 2θ zero shift can be included when there are enough indexed reflections. The search is restricted to a user-defined ± range. This can compensate for a simple global offset, but it is not a substitute for a complete instrument calibration or correction of specimen-displacement and other geometry errors.

How many indexed reflections are needed?

The minimum number depends on the crystal system because different systems contain different numbers of independent metric parameters. A formally solvable least-squares system is only the starting point; practical validation usually requires more reliable independent reflections than the mathematical minimum.

Observed vs Calculated Reflections

For each accepted hkl assignment, ChemFitX reports observed 2θ, calculated 2θ, Δ2θ, observed d-spacing and calculated d-spacing. Small residuals are useful evidence that the cell is internally consistent with the selected reflections, but residual size alone does not establish phase identity or chemical correctness.

Calculated Reflection Positions

After refinement, the platform can generate calculated reflection positions within the experimental 2θ range using the refined cell and selected centering rule. These are reflection positions only. The current module does not calculate full structure-factor intensities, preferred-orientation effects, profile shapes or a complete powder-pattern simulation.

Optional Reference CIF Workflow

Researchers who already know the likely phase can load a crystallographic information file (CIF) as an advanced reference. The CIF is parsed locally to obtain a complete unit cell and, when present, space-group metadata, chemical formula, Z and the number of fractional atom sites.

Use CIF Cell

Use CIF Cell transfers the reference lattice parameters into the platform and infers a compatible crystal-system setting. When the CIF contains chemical formula and Z, those fields are also transferred to the theoretical-density section.

Match Peaks to CIF

Match Peaks to CIF generates reflection positions from the reference cell and assigns tentative hkl values to observed peaks that fall within the selected angular tolerance. The assignments can then be reviewed and refined against the experimental peak positions.

This reference matching is cell-guided rather than a full structure-factor simulation. The current workflow does not apply every space-group extinction condition and does not calculate diffraction intensities from atomic coordinates, so a matched peak list should still be validated crystallographically.

Starting-Cell hkl Suggestions

If an approximate unit cell is already known but a reference CIF is not available, the advanced starting-cell tool can suggest hkl values from user-entered lattice parameters. It can also use a selected centering filter and configurable index and tolerance limits.

This assisted mode is different from Auto Index & Refine: the starting-cell tool depends on the supplied approximate cell, whereas automatic indexing searches candidate cells without requiring one. Full space-group extinction conditions are not inferred automatically in the starting-cell workflow.

Theoretical Density Calculation

The theoretical-density module requires a refined unit-cell volume, chemical formula and Z, the number of formula units per unit cell. If a loaded CIF contains formula and Z, ChemFitX can fill those fields automatically; otherwise they can be entered manually.

The platform first calculates formula molar mass from the entered chemical formula and then uses the refined cell volume to report density in g/cm³. Density cannot be determined from XRD peak positions alone because the chemical composition and formula-unit multiplicity are essential inputs.

Interpreting Candidate Scores and Residuals

A high candidate score or a low RMS Δ2θ indicates that a proposed cell can reproduce the selected peak positions relatively well within the implemented search model. It does not prove that the cell is chemically plausible, uniquely correct or associated with the intended phase.

When judging an indexed cell, consider the fraction of reliable peaks indexed, residual distribution, systematic unindexed peaks, expected lattice dimensions, known chemistry, possible multiphase contributions, reference data and independent characterization. A solution that fits only a subset of strong peaks can still be misleading.

hkl Labels and Graph Presentation

Indexed peaks can be displayed with hkl labels and leader arrows on the XRD pattern. Font size, text color, arrow color and label background can be customized. Labels can also be dragged to clearer positions without changing the underlying peak assignments.

Exporting Crystallographic Results

Indexed CSV

The indexed CSV contains observed peak number, 2θ, intensity, observed d-spacing, h, k, l, inclusion status and assignment method. This is useful for documenting the exact peak set used during interpretation.

Refined-cell CIF

The refined-cell CIF contains the refined unit-cell metrics and observed indexed reflection metadata. It does not contain atomic coordinates and is not a solved crystal structure. It should therefore be treated as a cell/reflection export rather than a complete structural CIF.

Calculated reflection list

The reflection-list export contains calculated h, k, l, 2θ and d-spacing values generated from the refined cell within the relevant angular range. Calculated intensities are not included.

Data Privacy and Browser-Based Processing

The XRD pattern and optional CIF are read and processed locally in the browser by this application. The files are not uploaded to ChemFitX servers during the crystallographic analysis workflow.

Researchers should still follow their laboratory or institution’s requirements for unpublished diffraction data, proprietary crystal structures, shared computers and exported files.

Scientific Scope and Limitations

The platform performs peak-position-based crystallographic indexing and unit-cell refinement; it is not a Rietveld-refinement engine.
Automatic indexing produces candidate cells and tentative hkl assignments. Candidate ranking is not proof of unique phase identification.
Peak detection can miss weak reflections or include artifacts, shoulders or overlapping maxima. Review the peak list before indexing.
Smoothing and baseline correction can shift, suppress or merge weak diffraction features when used excessively.
Accurate d-spacing requires the correct X-ray wavelength and a correctly calibrated 2θ axis.
A refined constant zero shift represents only a simple global angular offset and does not model every instrumental or specimen-related error.
Simple lattice-centering filters are supported, but full space-group systematic absences from screw axes and glide planes are not automatically enforced.
Calculated reflection tables contain positions, not full structure-factor intensities or a simulated powder pattern.
Reference-CIF matching uses the reference cell for peak-position guidance and does not by itself constitute full structural refinement.
The exported refined-cell CIF contains cell metrics and reflection information but no solved atomic coordinates.
Theoretical density requires the refined cell volume, chemical formula and Z; it cannot be obtained from peak positions alone.
Multiphase patterns, severe peak overlap, preferred orientation, poor peak calibration and missing low-angle reflections can make indexing ambiguous or incorrect.
Final crystallographic conclusions should be checked against chemical plausibility, reference diffraction data and complementary characterization where appropriate.

Frequently Asked Questions

What does the ChemFitX Crystallographic Analysis Platform analyze?

It analyzes powder XRD peak positions for d-spacing, candidate hkl indexing and indexed unit-cell refinement. It also provides optional reference-CIF matching and theoretical-density calculation.

Which XRD file formats can I upload?

The platform accepts CSV, TSV, TXT, XY, XYE, ASC, text DAT and supported XRDML scans.

Can ChemFitX automatically index an XRD pattern?

It can search candidate unit cells, propose tentative hkl assignments and refine candidate solutions from detected peak positions. The results should be reviewed because automatic powder indexing can be non-unique.

Do I need to know the crystal system before indexing?

No. The Auto setting searches candidate models across all seven supported crystal systems. If the system is independently known, you can restrict the search.

Which crystal systems are supported?

Cubic, tetragonal, orthorhombic, hexagonal, trigonal/rhombohedral, monoclinic and triclinic systems are supported for indexed unit-cell refinement.

Is this a Rietveld refinement tool?

No. The crystallographic module refines unit-cell metrics from indexed peak positions. It does not refine a full calculated powder pattern against observed intensities as in Rietveld refinement.

Can I load a CIF file?

Yes. A reference CIF can provide a known unit cell and metadata such as space group, chemical formula and Z when those fields are present. The reference cell can be used for peak matching and refinement guidance.

Does CIF matching use atomic coordinates to calculate full intensities?

No. The current reference-CIF workflow is primarily cell-guided. It does not perform full structure-factor intensity calculation.

Can the platform identify the space group automatically?

Not from the current peak-position workflow alone. Simple centering filters and CIF space-group metadata can be used, but complete space-group extinction analysis is outside this module.

What does RMS Δ2θ mean?

RMS Δ2θ summarizes the angular differences between observed indexed peaks and the positions calculated from the refined unit cell. Lower residuals indicate a better positional fit for the selected assignments, but do not alone prove that the cell is correct.

Can I refine zero shift?

Yes. A constant 2θ zero shift can optionally be refined within a user-defined range when the indexed dataset contains enough information.

Can ChemFitX calculate theoretical density?

Yes, after unit-cell refinement. The calculation also requires a chemical formula and Z. These may be imported from a reference CIF when available or entered manually.

What is exported in the CIF file?

The exported CIF contains the refined cell metrics and observed indexed reflection metadata. It does not contain atomic coordinates and is not a solved structural model.

Are my XRD and CIF files uploaded to ChemFitX?

For this application, the files are read and analyzed locally in the browser and are not uploaded to ChemFitX servers by the analysis code.

Scientific References

Cullity, B. D.; Stock, S. R. Elements of X-Ray Diffraction. Pearson.
Pecharsky, V. K.; Zavalij, P. Y. Fundamentals of Powder Diffraction and Structural Characterization of Materials. Springer.
International Tables for Crystallography, Volume A: Space-Group Symmetry. International Union of Crystallography.