quality control raises a handful of sensible questions. This page answers them in order, starting with the fundamentals and moving to applications.
This page was last updated on 2025-09-25 and is reviewed periodically as new material appears.
High-performance liquid chromatography, or HPLC, separates dissolved compounds by passing a liquid mobile phase through a packed column. Components distribute differently between the stationary phase and the moving liquid, so they travel at different speeds and exit at different times. A detector records these eluting bands as peaks, and peak area or height relates to amount. The technique supports testing in pharmaceuticals, foods, environmental samples, and industrial chemicals. Quantification usually depends on calibration with known standards.
Several separation modes exist, including reversed-phase, normal-phase, ion-exchange, size-exclusion, and hydrophilic interaction liquid chromatography. Reversed-phase uses a nonpolar stationary phase with a polar mobile phase and is widely applied to small organic molecules. Gradient elution changes mobile phase composition during the run, while isocratic elution keeps it constant. Column chemistry, particle size, temperature, flow rate, and mobile phase pH all influence retention and resolution. Method development selects conditions that separate analytes from matrix components and from each other.
Method validation establishes that an HPLC procedure is suitable for its intended use. Key parameters include accuracy, precision, specificity, linearity, range, limit of detection, limit of quantitation, and robustness. Accuracy measures agreement with a true or accepted value, while precision describes repeatability and intermediate precision. Specificity confirms that the method measures the analyte without interference from impurities, degradants, or excipients. Validation is documented in a protocol and report, and acceptance criteria are set before experiments begin. Regulatory guidance varies by region, but the general principles are widely harmonized.
System suitability testing is performed before and during analytical runs to confirm that the instrument and method are working as expected. Common checks include retention time, peak area, resolution between critical pairs, tailing factor, and theoretical plate count. Results are compared with predefined limits, and a failed check requires investigation before sample results are reported. Quality control samples at low, middle, and high concentrations are injected at intervals to monitor accuracy and precision. Blank injections detect carryover and contamination, while control charts track performance over time.
Data handling and documentation are central to HPLC quality control. Electronic systems should have audit trails that record changes to methods, sequences, and results. Integration parameters, such as peak baseline and threshold, can affect reported areas and must be defined in advance. Out-of-specification results trigger a structured investigation that may include reanalysis, instrument checks, and review of sample preparation. Regulatory inspections often examine raw data, audit trails, and training records to verify that reported results are traceable and reliable.
| Property | Value | Notes |
|---|---|---|
| Column particle size | 3–5 µm for conventional HPLC; sub-2 µm for UHPLC | Smaller particles increase backpressure and efficiency. |
| Typical flow rate | 0.5–2.0 mL/min for a 4.6 mm internal diameter column | Flow scales with column diameter and particle size. |
| UV detection wavelength | 190–400 nm | Selection depends on analyte chromophore. |
| Column temperature | 25–40 °C | Temperature affects retention, selectivity, and pressure. |
| Injection volume | 1–20 µL | Larger volumes may distort early-eluting peaks. |
Validation establishes that a method is suitable for its intended purpose. Typical parameters include accuracy, precision, specificity, linearity, range, limit of detection, limit of quantification, robustness, and stability of standards and samples. Acceptance criteria are defined in advance, and results are documented in a validation report. Regulatory guidance for pharmaceuticals, foods, and environmental testing differs, so the applicable framework must be identified. Ongoing verification uses control samples and trend charts after validation. Method transfer to another laboratory may require partial revalidation.
Routine quality control includes blanks, duplicates, spiked samples, and certified reference materials. Calibration curves are prepared with standards at several concentrations, and the detector response is checked for linearity. Carryover, column aging, mobile phase evaporation, and temperature drift can shift retention times or peak areas. Maintenance such as replacing seals, filters, and columns helps prevent failures. Records of injections, integration, and deviations support traceability. Audits may request raw data and instrument logs for each batch.
Documentation and traceability are central to regulated HPLC testing. Records typically include instrument logs, column history, mobile-phase preparation, sample preparation, injection sequences, raw chromatograms, and audit trails. Electronic systems may require user access controls, time-stamped changes, and backup procedures. Training records show that analysts are qualified for assigned methods. Audits and inspections check whether written procedures match actual practice and whether deviations are documented. These controls support reproducibility and allow results to be reconstructed if questions arise later.
Method validation establishes that an HPLC procedure is suitable for its intended purpose. Typical parameters include accuracy, precision, specificity, linearity, range, limit of detection, limit of quantitation, robustness, and solution stability. Accuracy reflects closeness to a reference value, while precision reflects agreement among repeated measurements. Specificity shows whether the method can measure the analyte without interference from matrix components. Validation is documented through protocols and reports, and the required extent depends on the method's use and regulatory context.
Routine quality control uses system suitability, blank injections, check standards, and control samples to detect drift or contamination. System suitability criteria may specify minimum resolution, maximum tailing factor, and a permitted range for repeated injections. Blank injections reveal carryover or solvent contamination, while check standards confirm calibration accuracy over a batch. Control samples with known analyte levels can show whether results remain within statistical limits. When a control result falls outside limits, the analyst investigates the cause and may invalidate affected results before repeating the batch.
Most routine HPLC testing uses reversed-phase columns, where the stationary phase is nonpolar and the mobile phase is a polar mixture such as water with an organic solvent. Analytes partition between the two phases according to polarity, size, and charge. Gradients that change solvent composition over time can separate compounds with broad retention ranges. Isocratic conditions keep solvent composition constant and suit simpler mixtures. The choice of column chemistry, pH, and temperature affects selectivity and peak shape.
Detection in HPLC testing commonly relies on ultraviolet-visible absorbance, fluorescence, refractive index, or mass spectrometry. UV detection is widely used because many organic compounds absorb light, but it requires a chromophore. Mass spectrometry provides mass-based identification and high sensitivity for trace analytes. Each detector has trade-offs in selectivity, cost, and compatibility with mobile phases. Quantification typically uses calibration curves prepared from reference standards. Results are reported as concentration, purity, or presence above a limit.
Pickled tofu (豆腐乳 in Chinese, pinyin: dòufurǔ, or 腐乳 fŭrŭ; chao in Vietnamese), also called "preserved tofu" or "fermented tofu", consists of cubes of dried tofu that have been allowed to fully air dry under hay and slowly ferment with the help of aerial bacteria. The dry fermented tofu is then soaked in salt water, Chinese rice wine, vinegar or minced chili peppers, or in a mixture of whole rice, bean paste, and soybeans. In the case of red pickled tofu (紅豆腐乳 in Chinese, Pinyin: hóng dòufurǔ), red yeast rice (cultivated with Monascus purpureus) is added for color. In Japan, pickled tofu with miso paste is called tofu no misodzuke, and is a traditional preserved food in Kumamoto. In the Ryukyu Islands, pickled and fermented tofu is called tōfuyō (豆腐餻). It is made from Shima-dōfu (an Okinawan variety of large and firm tofu). It is fermented and matured with koji mold, red koji mold, and awamori. Stinky tofu (臭豆腐 in Chinese, Pinyin: chòudòufu) is soft tofu that has been fermented in a vegetable and fish brine. The blocks of tofu have a pungent cheese smell, sometimes resembling rotting food. Despite its strong odor, the flavor and texture of stinky tofu is appreciated by aficionados, who describe it as delightful. The texture of this tofu is similar to the soft East Asian tofu from which it is made. The rind that stinky tofu develops when fried is said to be best when especially crisp, and fried stinky tofu is usually served with soy sauce, sweet sauce, or hot sauce.
== Lawrence Livermore Laboratory == In 1957, as a captain in the United States Air Force, Robert Finnigan joined the University of California Radiation Laboratory at Livermore (later the Lawrence Livermore National Laboratory). Finnigan worked on Project Pluto, a United States government program to develop nuclear powered ramjet engines for the Supersonic Low Altitude Missile (SLAM), a type of cruise missile. Finnigan led the group that developed computer control systems for the nuclear reactor that powered the missile. In 1959, Finnigan had the choice of rotating to a different tour of duty in the Air Force, or resigning from the Air Force and staying at Livermore. He remained at Livermore for two more years. Finnigan's group developed the Tory II-A prototype and Tory II-C reactor and controls, which were tested at the Nevada Test Site in the 1960s.
===== Mushroom burial ===== Mushroom burial has been developed by Jae Rhim Lee and her colleagues to address the impact traditional burial approaches have on the environment. It is an eco-friendly process which consists of dressing the cadaver in a bodysuit with mushroom spores woven into it, nicknamed the Infinity Burial Suit. Rhim developed her own mushrooms by feeding them her hair, skin, and nails to create a mushroom variety that will best decompose human remains. As the mushrooms grow, they consume the remains within the suit as well as the toxins that are being released by the body. Rhim and her colleagues created this suit as a symbol of a new way for people to think about the relationship between their body after death and the environment.
Sources: en.wikipedia.org
Categorical data can be either nominal or ordinal. Ordinal data has a ranked order for its values and can therefore be converted to numerical data through ordinal encoding. An example of ordinal data would be the ratings on a test ranging from A to F, which could be ranked using numbers from 6 to 1. Since there is no quantitative relationship between nominal variables' individual values, using ordinal encoding can potentially create a fictional ordinal relationship in the data. Therefore, one-hot encoding is often applied to nominal variables, in order to improve the performance of the algorithm. For each unique value in the original categorical column, a new column is created in this method. These dummy variables are then filled up with zeros and ones (1 meaning TRUE, 0 meaning FALSE). Because this process creates multiple new variables, it is prone to creating a 'big p' problem (too many predictors) if there are many unique values in the original column. Another downside of one-hot encoding is that it causes multicollinearity between the individual variables, which potentially reduces the model's accuracy. Also, if the categorical variable is an output variable, you may want to convert the values back into a categorical form in order to present them in your application. In practical usage, this transformation is often directly performed by a function that takes categorical data as an input and outputs the corresponding dummy variables. An example would be the dummyVars function of the Caret library in R.
Following months of rumours and investigations, on 13 August 2013, Reid—along with the Essendon Football Club, senior coach James Hird, senior assistant coach Mark Thompson, and football manager Danny Corcoran—was charged by the AFL with bringing the game into disrepute in relation to the supplements program at the club in 2011 and 2012. The club was given 14 days to consider the charges and face an AFL Commission hearing on 26 August 2013. On 27 August 2013, the AFL Commission handed down its decision to exclude Essendon from the 2013 finals series and fine it $2 million. James Hird was banned for 12 months, Danny Corcoran suspended for six months, and Mark Thompson fined $30,000, but the case against Reid continued after his decision to fight the charges against him. On 18 September 2013, 24 hours before Reid was due back in the Supreme Court, the AFL withdrew all 38 charges against him for his role in the supplements scandal, meaning he was free to resume his work at the Essendon Football Club.
PABA is an intermediate in the synthesis of folate by bacteria, plants, and fungi. Many bacteria, including those found in the human intestinal tract such as E. coli, generate PABA from chorismate by the combined action of the enzymes 4-amino-4-deoxychorismate synthase and 4-amino-4-deoxychorismate lyase. Plants produce PABA in their chloroplasts, and store it as a glucose ester (pABA-Glc) in their tissues. The malarial protozoan Plasmodium only make PABA when necessary, preferring to get it from the surroundings if able to. Some bacteria, including a few found in the human microbiome, are unable to make PABA for themselves but can use PABA to make folate. A few are very efficient at the PABA-to-folate conversion despite not making their own PABA. Sulfonamide drugs are structurally similar to PABA, and their antibacterial activity is due to their ability to interfere with the conversion of PABA to folate by the enzyme dihydropteroate synthetase. Thus, bacterial growth is limited through folate deficiency.
The three substrates of this enzyme are D-glyceraldehyde 3-phosphate, phosphate (Pi), and oxidised nicotinamide adenine dinucleotide (NAD+). Its products are 1,3-bisphosphoglyceric acid, reduced NADH, and a proton. The enzyme can use nicotinamide adenine dinucleotide phosphate as an alternative cofactor. This enzyme belongs to the family of oxidoreductases, specifically those acting on the aldehyde or oxo group of a donor with NAD+ or NADP+ as an acceptor. The systematic name of this enzyme class is D-glyceraldehyde 3-phosphate:NAD(P)+ oxidoreductase (phosphorylating). Other names in common use include triosephosphate dehydrogenase (NAD(P)) (phosphorylating), and glyceraldehyde-3-phosphate dehydrogenase (NAD(P)) (phosphorylating).
Sources: en.wikipedia.org
HPLC separates and quantifies compounds in a liquid sample. Detectors produce a response proportional to the amount of a compound passing through the flow cell. Identification by retention time requires comparison with a known standard.
UHPLC uses columns with smaller particles and operates at higher pressures than conventional HPLC. These conditions can improve speed, resolution, and sensitivity. Both techniques use the same fundamental separation principles.
Validation shows that a method performs reliably for its intended purpose across a defined range. It assesses accuracy, precision, specificity, linearity, and robustness. Regulated testing often requires documented validation before routine use.
System suitability is a set of checks that confirm the instrument and method perform within limits before sample analysis. It typically includes resolution, tailing factor, retention time, and peak area reproducibility. If a check fails, the run is invalidated until the cause is resolved.