¶LJPcalc Theory
¶Table of Contents
¶Calculation Method
LJPcalc calculates the liquid junction potential according to the stationary NernstPlanck equation which is typically regarded as superior to the simpler Henderson equation used by most commercial LJP calculators. Both equations produce nearly identical LJPs, but the Henderson equation becomes inaccurate as ion concentrations increase, and also when calculating LJP for solutions containing polyvalent ions.
¶LJPcalc Ion Mobility Library
LJPcalc uses an extensive ion mobility library: IonTable.md.html
The ion mobility table is stored in Markdown format. Not only does Markdown make it easy to display the table nicely in a browser,
but it also makes the table easy to edit in any text editor. Users desiring to use their own ion mobilities or add new ions to the table
can do so by editing the IonTable.md
file adjacent to LJPcalc.exe
as needed.
¶Influence of Ion Sequence on Calculated LJP
💡 LJPcalc automatically sorts the ion table into an ideal sequence prior to solving for LJP. Attention only needs to be paid to the ion sequence if automatic sorting is disabled.
When calculating LJP for a set of ions it is important to consider the sequence in which they are listed. Additional information can be found in Marino et al., 2014 which describes the exact computational methods employed by LJPcalc.

The last ion's c0 may be overridden to achieve electroneutrality on the c0 side. This will not occur if the sum of charge on the c0 side is zero.

cL for most ions will be slightly adjusted to achieve electroneutrality on the cL side. The secondtolast ion's cL (which cannot equal its c0) will remain fixed, while the last cL will be adjusted to achieve electroneutrality. During the solving process all cL values (but the secondfromlast) will be slightly adjusted. The adjustments are likely negligible experimentally, but this is why cL values in the output table slightly differ from those given for inputs.
¶How to Correct for LJP in Electrophysiology Experiments
Electrophysiologists often measure (or clamp) the voltage of cells impaled with glass microelectrodes, but the difference in ionic composition between the intracellular (pipette) and extracellular (bath) solutions produces a LJP which is added to the measurements:
V_{measured} = V_{cell} + LJP
To compensate for LJP, the electrophysiologist must calculate LJP mathematically (using software like LJPcalc) or estimate it experimentally (see the section on this topic below). Once the LJP is known, recorded data can be adjusted to accurately report cell voltage:
V_{cell} = V_{measured}  LJP
⚠️ This method assumes that (1) the amplifier voltage was zeroed at the start of the experiment when the pipette was in opentip configuration with the bath, and (2) the concentration of chloride (if using Ag/AgCl electrodes) in the internal and bath solutions are stable throughout the experiment.
¶Example PatchClamp LJP Calculation & Correction
This ion set came from in Figl et al., 2003 Page 8. They have been loaded into LJPcalc such that the pipette solution is c0 and the bath solution is cL. Note that the order of ions has been adjusted to place the most abundant two ions at the bottom. This is ideal for LJPcalc's analytical method.
Name  Charge  pipette (mM)  bath (mM) 

K  +1  145  2.8 
Na  +1  13  145 
Mg  +2  1  2 
Ca  +2  0  1 
HEPES  1  5  5 
Gluconate  1  145  0 
Cl  1  10  148.8 
Loading this table into LJPcalc produces the following output:
Values for cL were adjusted to achieve electroneutrality:
Name  Charge  Conductivity (E4)  C0 (mM)  CL (mM)

K  +1  73.5  145  2.8098265
Na  +1  50.11  13  144.9794365
Mg  +2  53.06  1  1.9998212
Ca  +2  59.5  0  0.9999109
HEPES  1  22.05  5  4.9990023
Gluconate  1  24.255  145  0
Cl  1  76.31  10  148.789725
Equations were solved in 88.91 ms
LJP at 20 C (293.15 K) = 16.052319631180264 mV
💡 Figl et al., 2003 Page 8 calculated a LJP of 15.6 mV for this ion set (720 µV lesser magnitude than our calculated LJP). As discussed above, differences in ion mobility table values and use of the NernstPlanck vs. Henderson equation can cause commercial software to report values slightly different than LJPcalc. Experimentally these small differences are negligible, but values produced by LJPcalc are assumed to be more accurate. See Marino et al., 2014 for discussion.
If we have patchclamp data that indicates a neuron rests at 48.13 mV, what is its true resting potential? Now that we know the LJP, we can subtract it from our measurement:
V_{cell} = V_{measured}  LJP
V_{cell} = 48.13  16.05 mV
V_{cell} = 64.18 mV
We now know our cell rests at 64.18 mV.
¶Notes about offset voltage, Ag/AgCl pellets, and and halfcell potentials
The patchclamp amplifier is typically zeroed at the start of every experiment when the patch pipette is in opentip configuration with the bath solution. An offset voltage (V_{offset}) is applied such that the V_{measured} is zero. This process nulls several potentials:
 liquid junction potential (caused by internal vs. bath solutions)
 halfcell potentials (caused by wire vs. internal and wire vs. bath)
 these potentials are large and variable when [Cl] is low on either side
 using an agar bridge helps keep this constant
When the amplifier is nulled prior to experiments the halfcell potentials can typically be ignored. However, if the [Cl] of the internal or bath solutions change during the course of an experiment (most likely to occur when an Ag/AgCl pellet is immersed in a flowing bath solution), the halfcell potentials become significant and affect V_{measured} as they change. See Figl et al., 2003 for more information about LJPs as they relate to electrophysiological experiments.
¶Measuring LJP Experimentally
It is possible to measure LJP experimentally. However, this technique is discouraged because issues with KCl reference electrodes make it difficult to accurately measure LJP (Barry and Diamond, 1970). However, this technique can be used in cases when ion mobilities are not known:
To measure LJP of an intracellular vs. extracellular solution for wholecell patchclamp experiments:
 Fill the recording pipette with intracellular solution
 Fill the bath with the identical intracellular solution
 Use a highmobility bath reference electrode with 3M KCl
 you can use a pipette filled with 3M KCl
 you can use freshly cut 3M agar bridge
 do not use an Ag/AgCl pellet (see note below)
 In currentclamp (I=0) adjust V_{offset} so V_{measured} is 0 mV
 Change the bath from pipette solution to extracellular solution
 If using an agar bridge, replace it with a new one
 Note the measured voltage (it should be negative)
 The inverse of this is the LJP (it should be positive)
 Future recordings can be compensated: V_{cell} = V_{measured}  LJP
⚠️ Use of an Ag/AgCl pellet will not produce accurate results. This is because intracellular solution typically has a low [Cl]. Using a KCl reference is ideal because intracellular solution has high [K] and extracellular solution has high [Cl] so there is excellent mobility in all cases.
¶Effect of Temperature on LJP
The LJP is temperature dependent. There are two sources of temperaturedependent variation: the Einstein relation and the conductivity table. The former can be easily defined at calculation time, while the latter requires modifying conductances in the ion mobility table. These modifications typically have a small effect on the LJP, so standard temperature (25C) can be assumed for most applications.
The Einstein relation defines diffusion as D = µ * k * T
where:
D
is the diffusion coefficientµ
(mu) is ionic mobilityk
is the Boltzmann constant (1.380649e23 J / K)T
is temperature (K)
The ion conductivity table is temperaturespecific. Ion conductivity was measured experimentally and varies with temperature. The ion conductivity table here assumes standard temperature (25C), but ion conductivity values can be found for many ions at nonstandard temperatures. LJPcalc users desiring to perform LJP calculations at nonstandard temperatures are encouraged to build their own temperaturespecific ion tables.
¶Calculating Ionic Mobility from Charge and Conductivity
Ionic mobility is µ = Λ / (N * e² * z)
where:
µ
(mu) is ionic mobility (m² / V / sec)Λ
(Lambda) is molar conductivity (S * cm²/ mol)N
is the Avogadro constant (6.02214076e23 particles / mol)e
is the elementary charge (1.602176634e19 Coulombs)z
is the absolute value of the elementary charge of the ion
¶References
 Marino et al. (2014)  describes a computational method to calculate LJP according to the stationary NernstPlanck equation. The JAVA software described in this manuscript is opensource and now on GitHub (JLJP). Figure 1 directly compares LJP calculated by the NernstPlanck vs. Henderson equation.
 Perram and Stiles (2006)  A review of several methods used to calculate liquid junction potential. This manuscript provides excellent context for the history of LJP calculations and describes the advantages and limitations of each.
 Shinagawa (1980) "Invalidity of the Henderson diffusion equation shown by the exact solution of the NernstPlanck equations"  a manuscript which argues that the Henderson equation is inferior to solved NernstPlanckPoisson equations due to how it accounts for ion flux in the charged diffusion zone.
 Lin (2011) "The Poisson The PoissonNernstPlanck (PNP) system for ion transport (PNP) system for ion transport"  a PowerPoint presentation which reviews mathematical methods to calculate LJP with notes related to its application in measuring voltage across cell membranes.
 NernstPlanck equation (Wikipedia)
 Goldman Equation (Wikipedia)
 EGTA charge and pH  Empirical determination of EGTA charge state distribution as a function of pH.