Sample Preparation, Purification and Instrument Calibration

The instrument cannot fix what the sample preparation broke. If the analyte was thermally degraded during a Soxhlet extraction, the LC-MS/MS will quantitate the degradation product and call it the parent. If the wash step on the SPE cartridge stripped half the analyte along with the matrix lipids, the recovery is 50 percent and nobody knows because no one ran a spike. If the urine was diluted twenty-fold to fit the calibration range and the analyte fell below LOD, the report says "not detected" when the analyte was very much there. None of these failures show up in the chromatogram. They show up in cross-examination, when the defence expert opens the validation report and finds that the method was never qualified for the matrix the case sample arrived in.

This is the standard the courts apply, whether or not they name it explicitly. A presumptive screen can be sloppy and still useful as a flag. A confirmatory result cannot. Every step from the moment the seal is broken on the evidence bag to the moment the analyte enters the source is part of the analytical method, and every step is open to challenge under Bharatiya Sakshya Adhiniyam Section 63 when the chemical examiner's certificate is tendered. The standard the courts apply, even when they do not name it, is the standard of method validation: was this workflow shown, in a documented validation study, to give an accurate answer for this matrix, this analyte and this concentration range? If yes, the result holds. If the laboratory cannot produce the validation file, the result is challengeable.

In a modern forensic toxicology workflow, the instrument is typically the smallest source of error. Sample preparation is the largest, and calibration sits between them. The extraction and clean-up steps warrant the most rigorous attention during method development and validation, because errors introduced before injection cannot be corrected by the instrument.

Extraction is the first move in sample preparation and the one that decides recovery. The technique is dictated by the matrix and the analyte's physical properties: polarity, volatility, thermal stability, the presence of protein binding and the concentration the analyst expects to see.

Liquid-liquid extraction (LLE) is the classical workhorse. The analyte partitions between an aqueous sample (blood, urine, viscera homogenate) and an immiscible organic solvent (dichloromethane, ethyl acetate, hexane, n-butyl chloride) chosen for its polarity match to the analyte. Acid-base manipulation drives the analyte into the right phase: alkalinise the aqueous phase to push basic drugs (morphine, cocaine, amphetamines) into the organic, acidify it to pull acidic drugs (barbiturates, salicylate) the same way. The classical Stas-Otto and ammonium-sulphate methods that Indian toxicology laboratories used for decades are LLE protocols with specific pH and salt conditions tuned to the analyte class. LLE is cheap, robust and works on almost any analyte that survives an organic solvent, but it is operator-intensive and produces emulsions on protein-rich matrices that need centrifugation to break.

Solid-phase extraction (SPE) is the modern toxicology workhorse. The sample passes through a small cartridge packed with a sorbent (Oasis HLB for mixed-mode reversed-phase, C18 for non-polar, MCX for strong cation exchange, MAX for strong anion exchange, Bond Elut Certify for drugs of abuse). The analyte adsorbs to the sorbent. Wash steps strip matrix interferents (salts, lipids, proteins) without dislodging the analyte. A final elution in a small volume of organic solvent releases the cleaned-up analyte into a tube ready for reconstitution and injection. SPE recoveries on validated methods sit comfortably in the 75 to 95 percent window for most drugs of abuse, the cleanup is dramatically better than LLE on biological matrices, and the cartridges are single-use to avoid carry-over. Every NABL-accredited Indian toxicology laboratory runs SPE as a routine part of the LC-MS/MS workflow.

Solid-phase microextraction (SPME) is the headspace and direct-immersion variant. A polymer-coated fused-silica fibre is exposed to the sample (or its headspace) and the analyte partitions onto the coating. The fibre is then thermally desorbed inside a GC injector. SPME is the technique of choice for fire debris (analysing accelerant residues), volatile organic compounds in urine (ethanol, methanol, ketones), and headspace screening of spiked beverages in drug-facilitated crime cases. The fibre needs no solvent, the cleanup is essentially zero and the sensitivity is excellent for volatiles.

Accelerated solvent extraction (ASE) and Soxhlet extraction sit at the other end. ASE uses elevated temperature and pressure to push solvent through a packed sample (soil, plant material, viscera) in minutes; Soxhlet uses refluxing solvent over hours. Both give exhaustive extraction of fat-soluble analytes (organochlorine pesticides, PCBs, polycyclic aromatic hydrocarbons in fire debris) where the analyte is bound to a complex matrix. Soxhlet is older, simpler and slow; ASE is faster, uses less solvent and is the modern choice for laboratories with a Dionex or equivalent instrument.

Dilute-and-shoot is the simplest workflow available. A urine sample is diluted ten- to fifty-fold with mobile phase, filtered through a 0.22 μm syringe filter, and injected directly. No extraction. The technique works only when the analyte concentration is well above LOQ in the diluted sample and when the matrix is simple (urine, dialysate). Modern LC-MS/MS sensitivity has made dilute-and-shoot a routine option for prescription pharmaceutical confirmation in urine at NDTL Delhi and several state SFSLs. The trade-off is matrix effect: ion suppression at the ESI source can be substantial because the matrix has not been cleaned up, and the deuterated internal standard becomes essential.

The extract that comes out of an SPE cartridge or a separating funnel is rarely ready to inject. Three more steps usually intervene: clean-up to strip residual interferents, derivatisation to make the analyte amenable to the instrument, and reconstitution to put the cleaned analyte into a solvent the column can handle.

Clean-up techniques are matrix-specific. Centrifugation at 10,000 to 14,000 g for ten minutes pellets protein precipitates after a methanol or acetonitrile crash, which is the simplest way to strip plasma proteins before LC injection. Filtration through a 0.22 or 0.45 μm PTFE syringe filter removes particulates that would otherwise clog the HPLC injector or guard column; PTFE is preferred over nylon for organic-rich extracts because nylon adsorbs basic drugs at higher pH. Dialysis through a 12 to 14 kDa cellulose membrane separates small-molecule analytes from large proteins and is the historic clean-up for drugs in whole blood when LC-MS/MS was not yet available. Florisil or silica column clean-up strips lipids and pigments from a fat-rich extract before GC-MS injection of pesticide residues. Gel permeation chromatography (GPC) on a Bio-Beads or equivalent column does the same job for polycyclic aromatic hydrocarbons in fire-debris extracts. Defatting with acetonitrile or n-hexane removes triglycerides from fatty viscera and adipose-rich tissue.

Derivatisation is the chemical step that converts a non-volatile, polar or thermally labile analyte into a form the GC will tolerate. The three workhorse derivatisation chemistries every Indian toxicology bench uses are silylation, methylation and acylation. Silylation with BSTFA (N,O-bis(trimethylsilyl)trifluoroacetamide) or MSTFA (N-methyl-N-(trimethylsilyl)trifluoroacetamide), often with one percent TMCS as a catalyst, replaces the active hydrogens on hydroxyl, amine and carboxylic acid groups with trimethylsilyl groups. The product is volatile, thermally stable and has a clean EI fragmentation pattern. BSTFA derivatisation is standard for opiates (morphine, codeine), benzoylecgonine (the cocaine metabolite), THC-COOH (the cannabis metabolite) and a long list of other drugs of abuse before GC-MS. Methylation with diazomethane or trimethylsilyldiazomethane converts carboxylic acids to methyl esters and is the routine pre-treatment for fatty-acid profiling and for some acidic drugs. Acylation with pentafluoropropionic anhydride (PFPA) or heptafluorobutyric anhydride (HFBA) gives heavy halogen-containing derivatives that improve sensitivity in negative chemical ionisation GC-MS, useful for low-concentration amphetamine and methamphetamine work.

Reconstitution is the final step before injection. The cleaned, derivatised extract is evaporated to dryness under a gentle nitrogen stream at 35 to 40 degrees Celsius (higher temperatures degrade thermally labile analytes), then redissolved in a small volume of a solvent compatible with the instrument: mobile phase for LC, the derivatising reagent or hexane for GC. The reconstitution volume sets the final concentration in the injection vial; a tenfold concentration over the original sample volume is typical for LC-MS/MS toxicology work. Mobile-phase compatibility matters more than people think. Reconstituting a reversed-phase LC sample in pure methanol gives a focusing problem on the column head and split or distorted peaks; reconstituting in 95 percent water with five percent methanol gives sharp peaks because the sample focuses tightly on the C18 stationary phase before the gradient pushes it off.

1. Spike the deuterated internal standard
   Add the IS at a known concentration to the original aliquot before any extraction step. d3-morphine for opiates, d5-EDDP for methadone, d5-alprazolam for triazolobenzodiazepines, pyrene-d10 for fire debris. The IS travels through every step from this point on.
2. Extract
   LLE, SPE, SPME, ASE or dilute-and-shoot, depending on matrix and analyte. The extraction parameters (solvent, pH, sorbent, flow rate, wash volumes, elution volume) come from the validated method, not from analyst memory.
3. Clean up
   Centrifuge, filter, dialyse or column-clean as the method requires. The aim is to remove matrix interferents that would compromise the chromatography or the ionisation.
4. Derivatise (if needed)
   BSTFA or MSTFA for silylation before GC, diazomethane for methylation, PFPA or HFBA for acylation. The derivatising reagent volumes and reaction temperature and time come from the method.
5. Evaporate and reconstitute
   Gentle nitrogen stream at 35-40 degrees Celsius to dryness, then redissolve in a known volume of a solvent compatible with the instrument's mobile phase or carrier.
6. Filter and transfer to vial
   0.22 μm PTFE syringe filter for LC; vial cap with PTFE-lined septum to avoid leaching. Label the vial with case number, sample ID and analyst initials. Inject.

A well-prepared sample injected into a poorly tuned instrument gives a poor result. The standardisation that comes before any case work is split into two distinct exercises: instrument calibration (tuning the instrument's hardware against a reference) and method calibration (building a concentration-response curve for the analyte in the matrix).

Instrument calibration is the hardware step. UV-Vis wavelength accuracy is checked against a holmium oxide filter or holmium oxide solution, both of which have well-defined absorbance peaks at known wavelengths (241.0, 287.5, 361.5, 451.4, 536.5 nm). FTIR wavenumber accuracy is checked against a polystyrene reference film with characteristic absorptions at 1601.4, 2849.5 and 3026.4 cm⁻¹; the test is part of the daily start-up at every NABL-accredited laboratory. Mass calibration on GC-MS is run on perfluorotributylamine (PFTBA, also called FC-43), a perfluorinated calibration gas with mass fragments at 69, 131, 219, 414, 502 and 614 m/z that span the working mass range. Mass calibration on LC-MS is run on sodium iodide for negative-mode tuning and on a vendor-specific ESI tune mix (Agilent ES Tuning Mix, Waters QC Reference Standard) for positive-mode tuning. Atomic absorption spectroscopy uses the hollow-cathode lamp profile for wavelength verification at the analyte's resonance line. ICP-MS uses a multi-element tune solution (typically Li, Y, Tl, Co at 1 ppb) to verify mass-axis linearity and resolution across the periodic table.

Method calibration is the analyst-facing step. A series of standards spanning the expected concentration range is prepared in the same matrix as the case sample (matrix-matched standards), spiked with the same internal standard at the same concentration, processed through the same workflow as the case samples (extraction, clean-up, derivatisation, reconstitution) and injected. The instrument response (peak area, absorbance, intensity) is plotted against the spiked concentration. Four calibration models cover almost every forensic application.

External calibration uses standards prepared in solvent rather than in matrix. Quick to prepare, fits well for clean samples (water, dilute organics), inadequate for biological matrices because matrix effects on extraction recovery and ionisation are not corrected. Internal-standard calibration uses the analyte-to-IS area ratio, plotted against analyte concentration. The IS corrects for losses during extraction and for matrix-effect ion suppression, which is why this is the default for all MS quantitation in forensic toxicology. Standard addition is the technique used when the matrix is heavy and the matrix effect cannot be corrected by an IS, typically metal quantitation in urine or whole blood by ICP-MS. The case sample is split into several aliquots; increasing known amounts of the analyte are spiked into each; the response is plotted against the added amount, and the unknown concentration is read where the line crosses zero on the negative x-axis (or by simple intercept calculation). Bracketing standards are not a curve type so much as a quality-control move: a calibration standard or QC sample is injected before, during and after the batch of case samples to track instrument drift through the run. The CFSL Chandigarh toxicology SOP requires bracketing every ten case samples; a failed bracket triggers re-run of every case sample between the last passing bracket and the failing one.

The acceptance criteria for the calibration curve itself are codified in NABL ISO 17025 method documents and in ICH Q2(R1) for pharmaceutical work. Linearity over the working range needs R² above 0.995. Each calibration point's back-calculated concentration must agree with its nominal value within ±15 percent (and ±20 percent at LOQ). The lowest calibrator sets the LOQ; the LOQ must be at or below the lowest concentration the case work is expected to encounter. The calibration curve must bracket the unknown; reading a case sample concentration above the highest calibrator or below the lowest is extrapolation and is not defensible at trial. If the case sample is too concentrated, dilute and re-run; if too dilute, concentrate and re-run.

Method validation is the documented exercise that proves the method works for the matrix and the analyte at the concentration range of interest. It is a one-time exercise per matrix and analyte pair, repeated when the method changes meaningfully. Without a current validation file, a forensic result cannot survive serious cross-examination on Section 63 BSA grounds. Nine parameters cover almost every method-validation requirement an Indian forensic laboratory faces.

Accuracy is measured as percentage recovery from a certified reference material or a spiked blank matrix. The accepted window for forensic toxicology is 80 to 120 percent of the nominal value across the working range; pharmaceutical analytical work tightens this to 95 to 105 percent at high concentrations. Precision is measured as relative standard deviation (RSD) of replicate measurements. Intra-day precision uses six replicates within one run; inter-day uses replicates over three to five days. The accepted threshold for forensic toxicology is RSD below 15 percent at all concentrations except LOQ, where RSD up to 20 percent is acceptable. Linearity is measured as R² of the calibration curve; the accepted threshold is above 0.995 over the full working range, with no systematic curvature that would suggest detector saturation or non-linear ionisation.

LOD is the lowest concentration the instrument can reliably distinguish from blank, defined as three times the standard deviation of blank (3σ method) or as the concentration giving a signal-to-noise ratio of 3:1. LOQ is where the response is reliable enough to put a number on, ten times SD of blank (10σ) or S/N of 10:1. Both numbers come from the validation study and both go on the case report.

Specificity (also called selectivity) is the demonstration that no interfering peak appears at the analyte's retention time and m/z transition in a blank matrix. A method that quantitates morphine but cannot resolve morphine from hydromorphone is not specific. The validation study includes a blank matrix run, a matrix spiked with the IS only, and a matrix spiked with structurally related compounds, all with no peak at the analyte's window. Robustness is the demonstration that small variations in conditions (mobile phase pH ±0.1, column temperature ±2 °C, flow rate ±5 percent) do not shift the result outside the precision window. Recovery is the spike-and-recover number measured in real matrix at three concentrations spanning the working range; the 80 to 120 percent window applies. Carry-over is the demonstration that a blank injection following the highest calibrator shows no analyte peak above 20 percent of the LOQ; if it does, the autosampler needle wash or the injection sequence needs revision.

A certified reference material is the anchor that ties an instrumental result to a national or international metrology standard. Without CRMs the laboratory's calibrators are themselves uncertified and the result is a number relative to nothing. Four classes of CRM cover almost every Indian forensic application.

NIST Standard Reference Materials are the international gold standard for inorganic analysis. NIST SRM 1577c bovine liver carries certified values for arsenic, lead, mercury, cadmium and a long list of essential and toxic elements; it is the standard CRM for ICP-MS toxic-metal panels in viscera and tissue at FSL Madhuban, CFSL Chandigarh and the AIIMS Forensic Medicine toxicology unit. NIST SRM 955c is lead in caprine blood at four certified concentration levels covering occupational and environmental exposure ranges; it is the reference for blood lead by graphite-furnace AAS or ICP-MS. NIST SRM 909 is human serum at certified concentrations for clinical chemistry analytes, used for method validation and quality control of LC-MS/MS toxicology methods that quantitate drugs in serum. The SRMs ship with a certificate of analysis listing the certified value, the expanded uncertainty and the traceability chain.

Indian Pharmacopoeia Reference Standards are produced by the Indian Pharmacopoeia Commission (IPC) at Ghaziabad. They cover prescription pharmaceuticals (paracetamol, diclofenac, alprazolam, diazepam, fluoxetine and several hundred others) at high purity with a certificate of analysis. IPC standards are the routine choice for drug-substance identification and quantitation work at SFSLs, CFSLs and the Central Drugs Standard Control Organisation (CDSCO) regional laboratories. They are cheaper than the NIST SRMs and are accepted by NABL as traceable references for pharmaceutical assays. Sigma-Aldrich (now Merck) certified solutions are the third class, sold as concentrated stocks of individual analytes (drugs of abuse, pesticides, environmental contaminants) traceable to NIST or to internal Sigma reference materials. They are the routine choice for drugs-of-abuse calibration solutions at toxicology benches.

The fourth class is in-house secondary standards, prepared from one of the certified primary standards above and used for daily calibration to conserve the primary CRM. The secondary standard's traceability is documented back to the primary CRM batch; the secondary's expiration is tracked and re-prepared when required.

NABL ISO 17025 imposes a quality-control schedule on every accredited laboratory. Daily QC requires at least one mid-range QC sample (a spiked blank matrix at a known concentration) injected with every batch; the result must fall within ±20 percent of the nominal value. Monthly CRM analysis requires a certified reference material to be analysed against the working calibration curve and the result reported into the laboratory quality system. Annual proficiency testing requires the laboratory to participate in an external scheme (the Indian Council of Medical Research toxicology PT scheme, or the German Society of Toxicological and Forensic Chemistry GTFCh scheme, or the Society of Forensic Toxicologists SOFT PT) where blind samples are analysed and results compared to consensus values.

The CFSL Chandigarh toxicology SOP, published in its quality manual and updated last in 2024, codifies bracketing standards every ten case samples, daily QC at low and high concentrations, monthly CRM analysis using NIST SRM 1577c and Sigma certified reference solutions, and annual GTFCh proficiency testing. FSL Madhuban (Karnal, Haryana) follows a comparable protocol with NIST SRM 955c as the blood-lead anchor. The CFSL and FSL practice underpins the chemical examiner's certificate that goes to the trial court under BSA Section 63; the certificate must reference the CRM batch number used for that day's calibration, and the defence is entitled to inspect that batch's certificate of analysis on application. A certificate that does not name the CRM batch is incomplete and is sometimes returned for re-issue by trial-court magistrates who know what to look for.