Technical Reference Guide

Alcohol Dilution:
The Mathematics and Physical Chemistry

A comprehensive guide to the equations, density data, and reasoning behind accurate dilution of ethanol and isopropanol — written for a reader with basic algebra and little prior chemistry knowledge. Covers both the table-based calculator and the IPA polynomial edition.

Published: 4 April 2026

1. Why Accurate Alcohol Dilution Matters

Ethanol and isopropanol are among the most widely used disinfectants and antiseptics in healthcare, pharmaceutical manufacturing, and laboratory settings worldwide. Both alcohols destroy bacteria and inactivate many viruses by denaturing proteins — disrupting the molecular machinery that microorganisms depend on to survive. However, this mechanism only works reliably within a specific concentration range. Diluting incorrectly — even by a few percent — can render a preparation clinically ineffective or unnecessarily wasteful of high-purity stock.

Accurate dilution also matters from a regulatory and safety standpoint. The World Health Organization (WHO) hand sanitizer formulations specify 80% v/v ethanol or 75% v/v isopropanol as their active concentrations.[1] These are not arbitrary numbers: they reflect decades of experimental evidence on antimicrobial efficacy balanced against skin tolerance and flammability.

Antimicrobial Efficacy Ranges

The U.S. Centers for Disease Control and Prevention (CDC) and WHO identify 60–90% v/v as the effective bactericidal range for both ethanol and isopropanol.[2] Below 50% v/v, efficacy drops sharply. Counterintuitively, concentrations above ~90–91% v/v are also less effective for routine disinfection — explained in the next subsection.

Property Ethanol Isopropanol
Minimum effective concentration~60% v/v~60% v/v
Optimum bactericidal range60–80% v/v60–90% v/v
WHO recommended formulation80% v/v75% v/v
USP rubbing alcohol specification—68–72% v/v
Activity vs. enveloped viruses (e.g. influenza, SARS-CoV-2)60–80% v/v — potent60–90% v/v — potent
Activity vs. non-enveloped viruses (e.g. adenovirus, rhinovirus)Active at 60–80% v/vLimited activity
Bactericidal vs. propanols at same concentrationSlightly lowerSlightly higher
Skin toleranceBetterLower
Kills bacterial spores?NoNo

Table 1.1 — Comparative antimicrobial properties of ethanol and isopropanol. Sources: CDC[2], WHO[1], Sauerbrei (2020)[3].

An important distinction exists between the two alcohols in antiviral coverage. Ethanol at 60–80% v/v is a potent viricidal agent against both enveloped viruses (those with a lipid membrane, such as influenza, herpes, and SARS-CoV-2) and many non-enveloped viruses (such as adenovirus, rhinovirus, and rotavirus).[2] Isopropanol, while effective against enveloped viruses, has more limited and less consistent activity against non-enveloped viruses — making ethanol the preferred choice when broad viricidal coverage is required.

Neither alcohol kills bacterial spores at any concentration. Spore-forming bacteria such as Clostridioides difficile require sporicidal agents (e.g. bleach or hydrogen peroxide) for elimination.

Why Pure Alcohol Is Less Effective Than Diluted Alcohol

This is one of the most counterintuitive facts in disinfection science. A preparation of 70% ethanol or 70% isopropanol kills bacteria and inactivates viruses more reliably than a preparation of 99% or 100% alcohol. The reason lies in the mechanism of action: alcohol denatures proteins by disrupting their folded three-dimensional structure. This reaction requires water as a co-solvent. Water acts as a catalyst that accelerates protein denaturation and allows the alcohol to penetrate the cell membrane fully.[2]

Pure (anhydrous) alcohol dehydrates the outer surface of a bacterial cell so rapidly that it forms a protective protein shell before the alcohol can penetrate to the interior — a process called coagulation without penetration. The organism may survive. Diluted alcohol, by contrast, penetrates the entire cell before coagulating all proteins, ensuring cell death. Additionally, water slows evaporation, giving the alcohol more contact time on a surface.[4]

This is precisely why accurate dilution is not just about meeting a specification — it is about understanding that the "right" concentration is a physical optimum, not an arbitrary bureaucratic threshold.

2. Fundamental Concepts

Percent by Volume (% v/v)

The most common way to express alcohol concentration in pharmaceutical and medical contexts is percent by volume, written as % v/v. It answers the question: "Of every 100 mL of the final mixture, how many millilitres are pure alcohol?"

Definition — percent by volume
\[\% \text{ v/v} = \frac{V_\text{alcohol}}{V_\text{mixture}} \times 100\]
where \(V_\text{alcohol}\) is the volume of pure alcohol and \(V_\text{mixture}\) is the total volume of the mixture, both measured at the same reference temperature.

The phrase "at the same reference temperature" is critical. The volume of a liquid changes with temperature — it expands when warmed and contracts when cooled. If you measure the alcohol volume at one temperature and the total volume at another, the resulting % v/v figure is meaningless. By convention, most pharmacopeial standards define % v/v at 20 °C, though 15 °C and 25 °C are also used.

Mass Fraction (% w/w)

An alternative way to express concentration is by mass. The mass fraction \(w\) is the ratio of the mass of alcohol to the total mass of the mixture:

Definition — mass fraction
\[w = \frac{m_\text{alcohol}}{m_\text{mixture}}\]
Dimensionless, ranging from 0 (pure water) to 1 (pure alcohol). Multiply by 100 to get percent by weight (% w/w).

Mass fraction has one extremely important property: it does not depend on temperature. If you have a solution with \(w = 0.70\) at 20 °C, it still has \(w = 0.70\) at 25 °C. The masses of the components have not changed; only the volume has changed (because of thermal expansion). This temperature-independence makes mass fraction the natural choice as an intermediate quantity in dilution calculations.

The Contraction Effect

Here is a surprising physical fact: if you mix 500 mL of pure ethanol with 500 mL of pure water, you do not get 1000 mL of mixture. You get approximately 970 mL. The mixture contracts. This is called the volume of mixing effect, and it is not a small correction — it is several percent and must be accounted for in any accurate dilution calculation.

Why does this happen? At a molecular level, ethanol and water are both polar molecules capable of forming hydrogen bonds. In the pure liquids, each molecule forms hydrogen bonds primarily with its own kind. When the two are mixed, ethanol and water molecules form hydrogen bonds with each other, and these cross-species bonds are energetically favourable. The molecules pack more tightly together in the mixture than they did separately, reducing the total volume. The effect is largest at intermediate concentrations (roughly 45–55% v/v) and approaches zero at the pure-component limits.

⚠ Important consequence

Because of the contraction effect, volumes are not additive when mixing ethanol and water. You cannot calculate how much water to add by working with volumes directly. Any correct dilution calculation must work through mass or mass fraction as an intermediate step.

Mass as the Bridge Quantity

Mass is additive — always. When you mix \(m_1\) grams of alcohol with \(m_2\) grams of water, you always get exactly \(m_1 + m_2\) grams of mixture, regardless of temperature or the contraction effect. This is simply conservation of mass. The calculator exploits this by using mass fraction as the central quantity that bridges all conversions:

1
Convert the input % v/v (at the reference temperature) to a mass fraction \(w\), using the mixture density.
2
Perform the dilution calculation entirely in grams using conservation of mass of the alcohol component.
3
Convert the resulting mass fraction back to % v/v (at either the reference temperature or the ambient temperature) using the mixture density again.

Steps 1 and 3 each require knowing the density of the mixture at a given concentration and temperature. This is the subject of the next section.

3. Density of Alcohol-Water Mixtures

Density \(\rho\) is the mass per unit volume of a substance, expressed here in grams per millilitre (g/mL) or equivalently in kg/m³ (1 g/mL = 1000 kg/m³). For alcohol-water mixtures, density is a function of both concentration and temperature. The two alcohols are handled differently because highly accurate equation-based data exists for ethanol but not for isopropanol at all temperatures.

Ethanol: The Wagenbreth-Blanke Polynomial (OIML R22)

For ethanol-water mixtures, the calculator uses a 54-coefficient polynomial equation developed by H. Wagenbreth and W. Blanke at the Physikalisch-Technische Bundesanstalt (PTB) in Germany, published in 1971 and adopted as the international standard in OIML Recommendation R 22 (1975).[5],[6] This equation gives density directly from the ethanol mass fraction and temperature, to within the precision limits of the underlying experimental measurements.

Background: How was the equation derived?

The Wagenbreth-Blanke polynomial is a least-squares fit to precision density measurements of ethanol-water mixtures taken over a grid of concentrations and temperatures. The form of the polynomial — separating pure-concentration terms (A), pure-temperature terms (B), and cross-interaction terms (C) — was chosen to provide an accurate, compact representation across the full range of concentration (0–100% by mass) and temperature (−20 °C to +40 °C). The coefficients were determined by fitting to the experimental data and were adopted without modification by OIML as the international alcoholometric reference. The original paper is: Wagenbreth, H. and Blanke, W. (1971), PTB-Mitteilungen 81, pp. 412–415. This publication is not freely available online; the coefficients used here are taken directly from the OIML R22 (1975) document, which is publicly available.

The Equation

The density \(\rho\) of an ethanol-water mixture in kg/m³ is given by:

Wagenbreth-Blanke Equation (OIML R22, 1975)
\[ \rho(p,\,t) = A_1 + \sum_{k=2}^{12} A_k\, p^{\,k-1} + \sum_{k=1}^{6} B_k\,(t-20)^k + \sum_{i=1}^{5}\sum_{k=1}^{m_i} C_{i,k}\, p^k\,(t-20)^i \]
Valid for \(-20\,°\text{C} \leq t \leq 40\,°\text{C}\) and \(0 \leq p \leq 1\).

Variable definitions:

\(\rho\)
Density of the ethanol-water mixture, in kg/m³. Divide by 1000 to convert to g/mL.
\(p\)
Mass fraction of ethanol — the ratio of the mass of ethanol to the total mass of the mixture. Dimensionless, ranging from 0 (pure water) to 1 (pure ethanol).
\(t\)
Temperature in degrees Celsius (°C).
\((t - 20)\)
Temperature deviation from the reference point of 20 °C. All thermal terms in the equation are expressed relative to 20 °C.
\(A_k\)
12 coefficients describing the concentration dependence of density at exactly 20 °C. Units: kg/m³ (since \(p\) is dimensionless). Note that \(A_1 = 998.20123\) kg/m³ is the density of pure water at 20 °C — the baseline value for the entire polynomial.
\(B_k\)
6 coefficients describing how the density of the pure-water baseline changes with temperature. Units: kg/(m³·°Ck).
\(C_{i,k}\)
36 interaction (cross) coefficients describing how the effect of concentration changes with temperature. Units: kg/(m³·°Ci).
\(m_i\)
Truncation limits for the C matrix: \(m_1=11,\; m_2=10,\; m_3=9,\; m_4=4,\; m_5=2\). Coefficients beyond these limits are exactly zero and are omitted.

Understanding the three groups of terms:

  • The A terms capture how density varies purely with ethanol content at exactly 20 °C — a polynomial in \(p\) alone.
  • The B terms capture how density changes purely with temperature when there is no ethanol (\(p=0\), pure water) — a polynomial in \((t-20)\) alone.
  • The C terms capture the interaction between concentration and temperature. They are necessary because the thermal behaviour of the mixture is not a simple combination of the pure-component behaviours — it depends on both \(p\) and \(t\) simultaneously.

Coefficients

The following table gives all 54 non-zero coefficients at full double-precision, taken directly from the OIML R22 (1975) PDF.[5] The C coefficient matrix has 5 rows and up to 12 columns (60 cells), but only 36 are non-zero — the remainder are exactly zero and omitted per the truncation limits \(m_i\). Together with 12 A coefficients and 6 B coefficients, the total is 54.

k Ak (kg/m³) Bk (kg/m³·°Ck) C1,k C2,k C3,k C4,k C5,k
1998.20123−0.206185130.169344346153009−0.0119301300505701−0.000680299573350384.07537667562203×10⁻⁶−2.78807435478241×10⁻⁸
2−192.9769495−0.0052682542−10.46914743455170.2517399633803460.0187683779028966−8.76305857347111×10⁻⁶1.34561288349335×10⁻⁸
3389.12389583.6130013×10⁻⁵71.9635346954652−2.17057570053699−0.2002561813734166.51503136009937×10⁻⁶—
4−1668.103923−3.8957702×10⁻⁷−704.74780542727913.53034988843031.02299296671922−1.51578483698721×10⁻⁶—
513522.154417.169354×10⁻⁹3924.09043003505−50.2998875854701−2.89569648390364——
6−88292.78388−9.9739231×10⁻¹¹−12101.6465906875109.6355666577574.81006058430068——
7306287.4042—22486.4655040079−142.275394642116−4.67214744079468——
8−613838.1234—−26055.6298218816108.0435942856232.45804310590346——
9747017.2998—18523.7392206947−44.1415323681739−0.541122762143681——
10−547846.1354—−7420.201433430147.44297153018878———
11223446.0334—1285.61784199897————
12−39032.85426——————

Table 3.1 — All 54 non-zero Wagenbreth-Blanke coefficients at full double-precision. "—" denotes a cell that is exactly zero and is omitted (see truncation limits \(m_i\)). The C matrix has 5 rows × 12 columns = 60 cells, of which 36 are non-zero. C column units are kg/(m³·°Ci) where i is the column's temperature-interaction order.

Source: OIML International Recommendation R 22 (1975), https://www.oiml.org/en/files/pdf_r/r022-e75.pdf

Ethanol Density Table

ℹ Computed values
All values in Table 3.2 are computed directly from the Wagenbreth-Blanke polynomial — they are not from a printed reference table. They are presented here for convenience. Because the polynomial is an empirical fit, computed values carry the accuracy of the underlying experimental data (approximately ±0.00005 g/mL).
% v/v 0 °C 15 °C 20 °C 25 °C 30 °C 35 °C 40 °C

Table 3.2 — Ethanol-water mixture density (g/mL) at 1% v/v resolution and seven temperatures. Computed from the Wagenbreth-Blanke polynomial (OIML R22, 1975). Reference temperature for the % v/v axis is 20 °C.

Isopropanol: Perry's Chemical Engineers' Handbook Tables

For isopropanol-water mixtures, no comparable polynomial of the same accuracy as Wagenbreth-Blanke has been adopted as an international standard. The calculator therefore uses empirical tabular data from Perry's Chemical Engineers' Handbook, 6th Edition (Green and Perry, 1984), page 3-92.[7] Perry's itself cites the International Critical Tables, Volume 3, page 120 as the primary source for this data.[8] The table values were accessed via handymath.com,[9] which reproduces the Perry's table in full and confirms the original citation.

The Perry's data is given in percent by weight (% w/w), not percent by volume. It covers temperatures of 0, 15, and 20 °C (with two independent observer datasets at 15 °C, noted in Perry's with * and **) and 30 °C. Values between tabulated concentrations are obtained by linear interpolation.

⚠ Two observer datasets at 15 °C

Perry's 6th Ed. presents two independent sets of density measurements at 15 °C for isopropanol-water mixtures, from two different observers. These are tabulated separately (15 °C* and 15 °C** in the table below). The two sets agree closely at the dilute and concentrated ends but differ by up to ~0.001 g/mL in the mid-range. The calculator uses the second dataset (15 °C**) for the 15 °C reference temperature. Both datasets are reproduced here for completeness.

Pure IPA Densities (CRC Handbook)

The density of pure isopropanol as a function of temperature is taken from the CRC Handbook of Chemistry and Physics, 95th Edition.[10] These values are used in the temperature extension method described below.

Temperature (°C)0152025303540
Pure IPA density (g/mL)0.801400.791300.785400.780940.776100.771300.76650

Table 3.3 — Density of pure isopropanol (2-propanol) at selected temperatures.

Source: CRC Handbook of Chemistry and Physics, 95th Ed., Haynes (ed.), 2014–2015. DOI: 10.1201/b17118

% w/w IPA0 °C15 °C *15 °C **20 °C30 °C

Table 3.4 — Isopropanol-water mixture density (g/mL) vs. percent by weight concentration, from Perry's Chemical Engineers' Handbook, 6th Ed., p. 3-92. Two independent observer sets are shown at 15 °C (* and **). Values at intermediate concentrations are obtained by linear interpolation between adjacent rows.

Primary source: Perry, R.H. and Green, D.W. (Eds.), Perry's Chemical Engineers' Handbook, 6th Ed., McGraw-Hill, 1984, p. 3-92, citing International Critical Tables, Vol. 3, p. 120 (1928). Data accessed via: handymath.com/calcshtml/IPAtable.html

⚠ Two suspected transcription errors corrected

Two values in this table violate strict monotonicity (density must decrease as % w/w increases at constant temperature) and are almost certainly digit transpositions from the original typesetting:

  • 13% w/w, 20 °C: Perry's (via handymath.com) prints 0.98760 — corrected here to 0.97860. The corrected value fits the 15°C→20°C drop pattern (drop ≈ 0.00140 g/mL, consistent with adjacent rows averaging 0.00127 g/mL). The printed value is 0.00900 g/mL above row 12, which is physically impossible.
  • 87% w/w, 30 °C: Perry's (via handymath.com) prints 0.82010 — corrected here to 0.81010. The corrected value fits between 86% (0.81270) and 88% (0.80780) with a step of 0.00260 g/mL per % w/w, consistent with surrounding values of 0.00240–0.00250.

These corrections have not been verified against the physical book. Anyone with access to Perry's 6th Ed. p. 3-92 or International Critical Tables Vol. 3, p. 120 is encouraged to confirm. In the standard calculator, linear interpolation between table rows means the corrected values only affect calculations near 13% and 87% w/w respectively.

Temperature Extension: 25 °C, 35 °C, and 40 °C

⚠ Extrapolated / interpolated values — not from Perry's

Perry's 6th Ed. provides isopropanol density data at 0, 15, 20, and 30 °C only. The calculator also accepts ambient temperatures of 25 °C, 35 °C, and 40 °C. Density values at these temperatures are not from Perry's — they are estimated using a pure-component correction method described below. These estimated values should be treated with appropriate caution: they are physically reasonable but are not validated against independent experimental data at these temperatures.

Method for 25 °C (interpolation): The density at 25 °C is estimated by linear interpolation between the tabulated 20 °C and 30 °C values:

Linear interpolation at 25 °C
\[ \rho_\text{IPA}(c,\, 25°\text{C}) \approx \rho_\text{IPA}(c,\, 20°\text{C}) + \frac{1}{2}\bigl[\rho_\text{IPA}(c,\, 30°\text{C}) - \rho_\text{IPA}(c,\, 20°\text{C})\bigr] \]
where \(c\) is the isopropanol concentration (% w/w).

Method for 35 °C and 40 °C (extrapolation beyond Perry's range): A pure-component correction anchored to the 30 °C Perry's value is used. The key assumption is that the change in mixture density from 30 °C to temperature \(t\) can be estimated from the known thermal behaviour of the pure components (pure IPA and pure water), weighted by the mass fraction of IPA:

Pure-component correction (extrapolation)
\[ \rho_\text{IPA}(c,\, t) \approx \rho_\text{IPA}(c,\, 30°\text{C}) + w\bigl[\rho_\text{pure\,IPA}(t) - \rho_\text{pure\,IPA}(30°\text{C})\bigr] + (1-w)\bigl[\rho_\text{water}(t) - \rho_\text{water}(30°\text{C})\bigr] \]
where \(w\) is the IPA mass fraction derived from the concentration \(c\) using the 20 °C mixture density; \(\rho_\text{pure\,IPA}(t)\) comes from Table 3.3; \(\rho_\text{water}(t)\) comes from the Wagenbreth-Blanke polynomial evaluated at \(p=0\).

This correction neglects changes in the interaction (excess volume) between IPA and water as temperature varies beyond 30 °C. In practice, for the concentrations and temperature range involved (30–40 °C), the correction is physically reasonable, but the estimated values carry additional uncertainty compared to the directly tabulated values.

Isopropanol — Alternative: Bivariate Polynomial (IPA Polynomial Edition)

ℹ IPA Polynomial Edition only

This section describes the isopropanol density method used in the IPA Polynomial Edition of the calculator. The standard calculator uses the Perry's table method described above. Both produce internally consistent and practically equivalent results; the differences are explained in the comparison section below.

Background and Motivation

The table-based method has one structural limitation: it requires access to the Perry's tabulated values, and those values only exist at discrete temperatures (0, 15, 20, 30 °C in Perry's directly, plus 25, 35, 40 °C by the correction method). Interpolation between table rows is linear, which is an approximation. An alternative approach is to fit a continuous polynomial surface \(\rho(p, t)\) to the same source data. Once the polynomial is fitted, density at any concentration and temperature within the valid range is obtained by direct evaluation — no table lookup or interpolation required.

No OIML-equivalent authoritative polynomial has been adopted for isopropanol-water mixtures. Unlike ethanol, where the Wagenbreth-Blanke equation is the international standard, the IPA polynomial used here was fitted by least squares specifically for this calculator, using the same Perry's source data as the table-based method. Its status is therefore that of a locally fitted approximation, not an officially adopted standard.

The Polynomial Equation

The IPA polynomial is a bivariate polynomial in mass fraction \(p\) and temperature \(t\), giving density in g/cm³:

IPA bivariate polynomial
\[ \rho(p,\,t) = \sum_{i=0}^{8} \sum_{j=0}^{4} C_{i,j}\; p^{\,i}\; t^{\,j} \]
Valid for \(0 \leq p \leq 1\) (mass fraction of IPA) and \(15\,°\text{C} \leq t \leq 40\,°\text{C}\). Result is in g/cm³ (= g/mL). The polynomial has 45 terms total (9 concentration powers × 5 temperature powers).

Variable definitions:

\(p\)
Mass fraction of isopropanol — the ratio of the mass of IPA to the total mass of the mixture. Dimensionless, ranging from 0 (pure water) to 1 (pure IPA). This is the same quantity as % w/w ÷ 100.
\(t\)
Temperature in degrees Celsius (°C). Valid range: 15–40 °C.
\(C_{i,j}\)
Fitted coefficients, one for each power combination \((p^i, t^j)\). Units: g·cm⁻³ (since \(p\) and \(t\) carry implied units in the fitting).
⚠ Valid range: 15–40 °C only

The polynomial was fitted to data that does not include Perry's 0 °C column, because the source data at 0 °C lies far outside the temperature range of practical use (15–40 °C) and would have caused the polynomial to extrapolate poorly within the normal working range. As a result, the IPA Polynomial Edition does not support a 0 °C reference or ambient temperature. For sub-15 °C applications, use the standard (table-based) calculator.

Coefficients

The 45 coefficients \(C_{i,j}\) are listed in the table below, in full double-precision as stored in the calculator. Each row corresponds to a power of \(p\) (\(i = 0\) to \(8\)), and each column to a power of \(t\) (\(j = 0\) to \(4\)).

\(i\) \\ \(j\) \(j=0\) \(j=1\) \(j=2\) \(j=3\) \(j=4\)
09.98625018679517917×10⁻¹1.24226188579390804×10⁻⁴−2.08242771960641887×10⁻⁶−3.22992012695510507×10⁻⁷4.98144660889320669×10⁻⁹
13.80724255873237105×10⁻¹−6.55253093282030658×10⁻²2.34964551379423592×10⁻³−2.29518648917080613×10⁻⁵−7.99093110980178926×10⁻⁸
2−4.736768089410204223.97949196718966614×10⁻¹3.21933188624022809×10⁻³−6.99218377653255388×10⁻⁴1.11552039790436701×10⁻⁵
31.18223787595893004×10¹1.42227529869033997−2.78239421931955611×10⁻¹1.13316161599800805×10⁻²−1.35928037481546895×10⁻⁴
4−9.15009327975162456−1.23123834873582005×10¹1.45501507479400094−5.26496863125275674×10⁻²5.99661957706848969×10⁻⁴
5−1.826646196936466903.00566682991122818×10¹−3.296934667267676031.16508944045049606×10⁻¹−1.31236773165147491×10⁻³
6−9.92844765767429194×10⁻¹−3.49960232301702305×10¹3.85576384766582780−1.36652351011857687×10⁻¹1.54219313290544801×10⁻³
71.18293034787813198×10¹1.96981716502191517×10¹−2.269231103005990858.17431666311367600×10⁻²−9.30222862941652135×10⁻⁴
8−7.50182170823811312−4.205878854660292995.28296280162725229×10⁻¹−1.95653213103377313×10⁻²2.25638935952954392×10⁻⁴

Table 3.5 — IPA bivariate polynomial coefficients \(C_{i,j}\) at full double-precision. \(i\) indexes powers of mass fraction \(p\); \(j\) indexes powers of temperature \(t\) (°C). Result is in g/cm³. Valid: 15–40 °C, \(0 \leq p \leq 1\).

Fitted by least squares to Perry's 6th Ed. (Table 3.4) at 15, 20, 30 °C (direct Perry's values), 25 °C (linear interpolation), and 35, 40 °C (pure-component extrapolation). Fit degree: \(p^8 \times t^4\) (45 terms). See Section 3.3.4 for fit statistics.

Accuracy and Limitations

The polynomial was fitted by ordinary least squares to the same Perry's source data that underlies the table-based method. Its fit quality was assessed against the data it was trained on:

MetricValue
Coefficient of determination (R²)0.9999946
Maximum residual (any point in fitting data)0.576 kg/m³ at 20 °C
Root-mean-square residual (RMS)0.147 kg/m³
Number of terms45 (degree 8 in \(p\), degree 4 in \(t\))
Valid temperature range15–40 °C
Valid concentration range0–100% w/w

Table 3.6 — IPA bivariate polynomial fit statistics.

A maximum residual of 0.576 kg/m³ (= 0.000576 g/mL) means the polynomial can deviate from the underlying Perry's table values by at most this amount at any given point. For typical pharmaceutical dilution calculations (target concentrations of 60–80% v/v), the corresponding error in the final calculated water volume is well under 1 gram per kilogram of concentrated alcohol — acceptable for practical laboratory use, but larger than the Wagenbreth-Blanke polynomial error for ethanol (which is limited only by the precision of the original measurements, approximately ±0.00005 g/mL).

📝 Source data provenance — identical to table method

The polynomial was fitted to the same Perry's 6th Ed. data described in Section 3.3 above, including the same temperature extensions at 25, 35, and 40 °C. All the caveats that apply to those extended temperatures in the table method apply equally to the polynomial — values at 25, 35, and 40 °C are interpolated or extrapolated from Perry's, not from independent measurements.

Comparison: Table-Based vs. Polynomial IPA Method

The two IPA density methods produce very similar practical results for typical dilution calculations. The table below summarises their key characteristics side by side.

Feature Table-based (standard calculator) Polynomial (IPA Polynomial Edition)
Underlying data source Perry's 6th Ed., p. 3-92 (via ICT Vol. 3, 1928) Same Perry's data
Temperature range supported 0–40 °C 15–40 °C only
Concentration interpolation Linear between 1% w/w rows Direct polynomial evaluation
Temperature interpolation Linear between tabulated temperatures Direct polynomial evaluation
Max density error vs. Perry's data Linear interpolation error only (~0.001 g/mL or less) 0.576 kg/m³ = 0.000576 g/mL (fit residual)
Official pedigree Yes — Perry's / ICT (widely cited) No — locally fitted; not independently validated
25, 35, 40 °C values Extrapolated/interpolated from Perry's Extrapolated/interpolated from Perry's (same source)
Recommended for All applications, especially <15 °C Convenience in 15–40 °C range; continuous evaluation

Table 3.7 — Comparison of the two IPA density methods. Both are derived from the same Perry's source data and produce equivalent results within the IPA polynomial's 15–40 °C range.

In summary: for most practical dilution work at room temperature or above, either method is appropriate and the results will agree to within the measurement precision of a laboratory balance. The polynomial has the small advantage of avoiding linear interpolation artefacts; the table-based method has the advantage of a wider temperature range and a direct, auditable connection to the printed Perry's values.

4. Converting Between % v/v and Mass Fraction

The dilution calculation requires converting an input % v/v concentration into a mass fraction, and then converting the resulting mass fraction back into % v/v. These conversions use density as the bridge.

Ethanol

The definition of % v/v connects volume fraction to mass fraction through density. At a reference temperature \(T_\text{ref}\), for a mixture with mass fraction \(w\) and density \(\rho_m\) (the mixture density), the volume of ethanol per unit total volume is:

% v/v from mass fraction — ethanol
\[ \% \text{ v/v} = \frac{w \cdot \rho_m(w,\,T_\text{ref})}{\rho_e(T_\text{ref})} \times 100 \]
where \(\rho_m(w,\,T_\text{ref})\) is the mixture density at mass fraction \(w\) and temperature \(T_\text{ref}\), and \(\rho_e(T_\text{ref})\) is the density of pure ethanol at \(T_\text{ref}\). Both densities are computed from the Wagenbreth-Blanke polynomial.

Going the other direction — converting a known % v/v to a mass fraction — requires solving the above equation for \(w\). Because \(\rho_m\) itself depends on \(w\) (through the Wagenbreth-Blanke polynomial), this is an implicit equation in \(w\) and cannot be solved with simple algebra. A numerical method is required.

Isopropanol

For isopropanol, the same relationship holds, but with the Perry's table used to look up \(\rho_m\) at a given temperature and (% w/w) concentration rather than the polynomial. Since Perry's data is in % w/w, the mass fraction is simply:

Mass fraction from % w/w — isopropanol
\[ w = \frac{\% \text{ w/w}}{100} \]

The complication for isopropanol is the inverse direction: given a % v/v, finding the % w/w. This also requires a Newton iteration, similar to the ethanol case.

Newton's Method — Making an Educated Guess and Refining It

Newton's method is an algorithm for finding the value of an unknown variable that satisfies an equation of the form \(f(w) = 0\). The idea is simple: start with a guess \(w_0\), compute how far off you are, and update the guess based on the slope of the function at that point.

Newton iteration
\[ w_{n+1} = w_n - \frac{f(w_n)}{f'(w_n)} \]
where \(f'(w_n)\) is the derivative of \(f\) evaluated at the current guess \(w_n\). The process repeats until \(|w_{n+1} - w_n|\) is smaller than a chosen tolerance (here, \(10^{-12}\)).

In this calculator, the function being solved for ethanol is:

Function to solve for % v/v → mass fraction
\[ f(w) = \frac{w \cdot \rho_m(w,\,T)}{\rho_e(T)} \times 100 - \% \text{ v/v}_\text{target} \]
The solution is the mass fraction \(w^*\) such that \(f(w^*) = 0\). Convergence typically occurs in fewer than 10 iterations for any point in the valid range.

5. The Dilution Calculation

Setting Up the Problem

The goal is to dilute a concentrated alcohol solution to a lower target concentration by adding water. We define the following quantities, all evaluated at the reference temperature \(T_\text{ref}\):

\(C\)
Concentration of the concentrated (starting) alcohol solution, in % v/v at \(T_\text{ref}\).
\(T\)
Target concentration of the diluted (final) solution, in % v/v at \(T_\text{ref}\).
\(\rho_A\)
Density of pure alcohol at \(T_\text{ref}\), in g/mL.
\(\rho_C\)
Density of the concentrated solution at concentration \(C\) and temperature \(T_\text{ref}\), in g/mL.
\(\rho_T\)
Density of the target (diluted) solution at concentration \(T\) and temperature \(T_\text{ref}\), in g/mL.
\(w_1\)
Mass fraction of alcohol in the concentrated solution.
\(w_2\)
Mass fraction of alcohol in the diluted solution.

The mass fractions are obtained from the % v/v concentrations using the density relationship:

Mass fractions from % v/v concentrations
\[ w_1 = \frac{C \cdot \rho_A}{100 \cdot \rho_C} \qquad\qquad w_2 = \frac{T \cdot \rho_A}{100 \cdot \rho_T} \]
These convert the volume-based concentrations \(C\) and \(T\) into mass-based concentrations \(w_1\) and \(w_2\), which are temperature-independent and conserved during mixing.

The key physical constraint is conservation of mass of the alcohol component: the amount of pure alcohol in the final diluted mixture equals the amount of pure alcohol that came from the concentrated solution. No alcohol is created or destroyed by adding water.

Mode A — Known Weight of Concentrated Alcohol

You know how much concentrated alcohol you have, and want to know how much water to add.

Mode A: Known \(m_C\) → find water to add

Given: \(m_C\) = mass of concentrated alcohol (grams)

Find: mass of water \(m_W\) to add

The mass of pure alcohol in the concentrated solution is:

\[\quad m_\text{alc} = m_C \cdot w_1\]

Since the diluted solution must contain the same mass of pure alcohol, and the diluted solution has mass fraction \(w_2\):

\[\quad m_F = \frac{m_\text{alc}}{w_2} = \frac{m_C \cdot w_1}{w_2}\]

The mass of water to add is the difference between the total final mass and the mass of concentrated alcohol added:

Water to add — Mode A
\[ \boxed{m_W = m_F - m_C = m_C\left(\frac{w_1}{w_2} - 1\right)} \]

Mode B — Known Weight of Final Diluted Product

You want to prepare a specific total mass of diluted solution and need to know how much concentrated alcohol and water to use.

Mode B: Known \(m_F\) → find concentrated alcohol and water needed

Given: \(m_F\) = total mass of diluted product (grams)

Find: mass of concentrated alcohol \(m_C\) and water \(m_W\)

The total mass of pure alcohol in the final product is:

\[\quad m_\text{alc} = m_F \cdot w_2\]

Since this alcohol comes entirely from the concentrated solution:

Concentrated alcohol needed — Mode B
\[ \boxed{m_C = \frac{m_\text{alc}}{w_1} = \frac{m_F \cdot w_2}{w_1}} \]

The water to add is:

\[\quad m_W = m_F - m_C\]
📝 Verification

The result can be verified by computing the mass fraction of the final mixture \(w_F = m_\text{alc} / m_F\), then converting back to % v/v at the reference temperature. If the calculation is correct, the resulting % v/v will equal the target \(T\) to within numerical precision.

6. Reference Temperature vs. Ambient Temperature

All the % v/v concentrations above are defined at the reference temperature \(T_\text{ref}\) — the temperature at which the original concentration specification is stated (commonly 20 °C). However, you may wish to make the solution at a different temperature — for example, in a warm laboratory at 30 °C. The calculator calls this the ambient temperature \(T_\text{amb}\).

The mass fraction of the diluted mixture \(w_F = m_\text{alc} / m_F\) is temperature-independent and is the same at all temperatures. However, the % v/v at ambient temperature \(T_\text{amb}\) will differ from the % v/v at \(T_\text{ref}\) because the mixture density changes with temperature.

To find the % v/v at ambient temperature, the calculator:

1
Takes the temperature-independent mass fraction \(w_F\) of the final diluted mixture.
2
Solves (by Newton iteration) for the % v/v at \(T_\text{amb}\) such that the resulting mass fraction matches \(w_F\).
3
Reports both the volume (using the mixture density at \(T_\text{amb}\)) and the % v/v at \(T_\text{amb}\).

If the reference and ambient temperatures are the same, this step is trivial and produces no change. If they differ, the % v/v at ambient will typically differ from the target by a fraction of a percent — the magnitude depends on how far apart the temperatures are and what the thermal expansion of the mixture is at that concentration.

ℹ Practical note

In pharmaceutical compounding, the specification (e.g. "70% v/v ethanol") is almost always stated at the reference temperature (commonly 20 °C). The physical mixing is typically performed at whatever the room temperature happens to be. The calculator allows you to specify both, so that you can correctly report the concentration either at the standard reference or at the actual preparation temperature.

7. Full Data Tables

The ethanol density table (Table 3.2) and the Perry's isopropanol table (Table 3.4) appear above in their respective sections. The IPA polynomial coefficient table (Table 3.5) and fit statistics (Table 3.6) appear in Section 3.4. This section collects pointers to all reference tables for convenience.

Ethanol Density (Wagenbreth-Blanke, OIML R22) — Table 3.2

Scroll up to Section 3.2 — Ethanol density table for the full 101-row table (0–100% v/v, seven temperatures). All values are computed from the polynomial and are not from a printed reference.

Isopropanol Density — Perry's 6th Ed. (Table 3.4)

Scroll up to Section 3.3 — Isopropanol for the full Perry's weight-concentration table. Note that the table is in % w/w (not % v/v), and that values at 25, 35, and 40 °C are extrapolated/interpolated and not from Perry's directly.

IPA Polynomial Coefficients (Table 3.5)

Scroll up to Section 3.4 — IPA bivariate polynomial for the 45-term coefficient table and fit statistics. Valid range: 15–40 °C. This method is used in the IPA Polynomial Edition of the calculator only.

8. Bibliography

  1. [1]
    World Health Organization. WHO Guidelines on Hand Hygiene in Health Care. Geneva: WHO Press, 2009. Includes the WHO hand rub formulations specifying 80% v/v ethanol or 75% v/v isopropanol. https://www.who.int/publications/i/item/9789241597906
  2. [2]
    Centers for Disease Control and Prevention (CDC). "Chemical Disinfectants." Guideline for Disinfection and Sterilization in Healthcare Facilities, 2008. Updated 2024. Primary source for bactericidal and viricidal concentration ranges. https://www.cdc.gov/infection-control/hcp/disinfection-sterilization/chemical-disinfectants.html
  3. [3]
    Sauerbrei, A. "Bactericidal and virucidal activity of ethanol and povidone-iodine." MicrobiologyOpen 9(9):e1097, 2020. Peer-reviewed review of concentration-dependent antimicrobial activity. https://pmc.ncbi.nlm.nih.gov/articles/PMC7520996/
  4. [4]
    Contec Inc. "Alcohols for Use as an Antimicrobial Agent." Technical reference article. Discussion of mechanism of action, contact time, and the role of water. https://cleanroom.contecinc.com/resource/alcohols-for-use-as-an-antimicrobial-agent
  5. [5]
    OIML (Organisation Internationale de Métrologie Légale). International Recommendation R 22: International Alcoholometric Tables. Paris: OIML, 1975. Primary source for the Wagenbreth-Blanke polynomial and all 54 non-zero coefficients used in this calculator. https://www.oiml.org/en/files/pdf_r/r022-e75.pdf
  6. [6]
    Wagenbreth, H. and Blanke, W. "Die Dichte des Wassers im Internationalen Einheitensystem und in der Internationalen Praktischen Temperaturskala von 1968." PTB-Mitteilungen 81, pp. 412–415, 1971. This PTB paper establishes the density of pure water on the 1968 International Practical Temperature Scale (IPTS-68), which provides the water-density baseline (the B-term series and the value A₁ = 998.20123 kg/m³) used in the Wagenbreth-Blanke polynomial. The full ethanol-water polynomial coefficients are given in OIML R22 [5]. Not freely available online.
  7. [7]
    Green, D.W. and Perry, R.H. (Eds.). Perry's Chemical Engineers' Handbook, 6th Edition. New York: McGraw-Hill, 1984, p. 3-92. Source of isopropanol-water density data at 0, 15, 20, and 30 °C (% w/w). Note: this edition is out of print; later editions (8th, 9th) are available from McGraw-Hill AccessEngineering. https://www.accessengineeringlibrary.com/content/book/9780071834087
  8. [8]
    Washburn, E.W. (Ed.). International Critical Tables of Numerical Data, Physics, Chemistry and Technology, Volume 3. New York: McGraw-Hill / National Research Council, 1928, p. 120. Cited by Perry's 6th Ed. as the primary source for the isopropanol density data.
  9. [9]
    Handymath.com. "Calculator for Making Isopropyl Alcohol in Water Solution." [Web page reproducing Perry's 6th Ed. Table, p. 3-92, with original citation.] Secondary source from which the Perry's table values were accessed and verified. https://www.handymath.com/calcshtml/IPAtable.html
  10. [10]
    Haynes, W.M. (Ed.). CRC Handbook of Chemistry and Physics, 95th Edition. Boca Raton: CRC Press / Taylor & Francis, 2014–2015. Source of pure isopropanol density values used in the temperature extension method. https://doi.org/10.1201/b17118
  11. [11]
    European Committee for Standardization (CEN). EN 14476:2019 — Chemical Disinfectants and Antiseptics: Quantitative Suspension Test for the Evaluation of Virucidal Activity in the Medical Area. Brussels: CEN, 2019. European standard for testing virucidal activity of disinfectants, referenced in the antimicrobial literature.

Alcohol Dilution: A Technical Guide — Version 1.2 — Last updated: 4 April 2026 — Changes: corrected W-B coefficient count (60→54; the C matrix has 60 cells but 36 are zero, giving 12+6+36=54 non-zero); integrated IPA polynomial section (§3.4–3.5); corrected Perry table errors at 13% w/w/20°C and 87% w/w/30°C; corrected OIML R22 title; fixed bibliography footnote anchors.