Beta Fulltext view is in preview — article structure may vary. Browse all articles
Contents
Medicinal & Analytical Chemistry International Journal Research Article 15 min read

On Choosing Ferrocene as an Internal Reference Redox Scale for Voltammetric Measurements: A Cautionary Tale

Angel AJ Torriero*
* Corresponding author
ISSN: 2639-2534  10.23880/macij-16000151  Received: November 25, 2019  Published: December 18, 2019
 — views
 1 references
 4 tables
PDF
Keywords
IRRS Stabilized Redox System Non-Aqueous
Abstract

Internal reference redox scales or internal reference redox systems, IRRS, are reversible or nearly reversible redox systems used to provide a known and stable reference point in non-aqueous solvents in conditions where reliable reference electrodes are difficult to be established and/or stabilized. In general, the IRRS is used in conjunction with quasi-reference electrodes. They have a long story, which began with the concept of Rb|Rb+ or Rb(Hg)|Rb+, followed by the use of organometallic redox couples. In order to limit the number of redox systems used as IRRS and then made easier the comparison, IUPAC recommended in 1983 that the systems ferrocene|ferrocenium, Fc0/+, and bis(biphenyl)chromium(0)|bis(biphenyl)chromium(I), BCr0/+, need to be used as internal reference redox systems in non-aqueous media. These two complexes were selected arbitrarily from several published redox systems.

Introduction

Internal reference redox scales or internal reference redox systems, IRRS, are reversible or nearly reversible redox systems used to provide a known and stable reference point in non-aqueous solvents in conditions where reliable reference electrodes are difficult to be established and/or stabilized [1, 2]. In general, the IRRS is used in conjunction with quasi-reference electrodes [3, 4, 5, 6, 7, 8]. They have a long story, which began with the concept of Rb|Rb⁺ or Rb(Hg)|Rb⁺, followed by the use of organometallic redox couples [9]. In order to limit the number of redox systems used as IRRS and then made easier the comparison, IUPAC recommended in 1983 that the systems ferrocene/ferrocenium, Fc⁰/*, and bis(biphenyl)chromium(0)|bis(biphenyl)chromium(I), BCr⁰/*, need to be used as internal reference redox systems in non-aqueous media. These two complexes were selected arbitrarily from several published redox systems.

An effective IRRS must possess a number of properties. Many of which were suggested by Gritzner and Kuta in their IUPAC recommendation [9], and most of which can be deduced by common sense. We recently suggested the following to be essential properties of an IRRS [10]:

- The OX_{IRRS}/RED_{IRRS} redox couple (OX_{IRRS} + e⁻ ⇌ RED_{IRRS}; being OX_{IRRS} and RED_{IRRS} the oxidised and reduced form of the IRRS, respectively) must be reversible or nearly reversible under the operative measurement conditions. It is advantageous for the IRRS electron transfer to be a simple one-electron outer-sphere electron transfer.

- It is very important that the IRRS redox moieties do not specifically interact with the working electrode (e.g., they do not adsorb on the working electrode).

- The mid-point potential for the IRRS should not overlap the active potential zone of the analyte (ANS) under study.

- The initially present IRRS redox component, OX_{IRRS} or RED_{IRRS}, must be stable over the time period of experimentation; it should not react (or interact) with the solvent, supporting electrolyte or ANS.

- The IRRS redox moiety, OX_{IRRS} or RED_{IRRS}, which is not initially present, must be stable on the time scale required to execute the slowest cyclic voltammetry (CV).

- The molecule or salt that is the source of the selected IRRS component should be pure, easily synthesized, or otherwise, readily available and should have a long shelf life.

There are also some constraints and requirements for the ANS [10]:

- The ANS is initially present as either an oxidizable moiety (RED_{ANS}) or reducible moiety (OX_{ANS}). It is worth noting that RED_{ANS} and OX_{ANS} species may or may not be redox partners, and the oxidation of RED_{ANS} and reduction of OX_{ANS} may be complicated.

- The initially present ANS moiety should be stable and should not react with the IRRS.

It is necessary to notice that IUPAC recommendation has been proposed for organic solvents and their mixtures. By extrapolation from those systems, the Fc⁰/* redox couple is also being used as an IRRS in ionic liquids, ILs [2, 11, 12, 13, 3].

Medicinal & Analytical Chemistry International Journal

Nowadays, it is well known that Fc is not an ‘ideal’ redox process as mentioned in the IUPAC recommendation, as its formal potential is highly dependent on the solvation effects of the organic solvent and supporting electrolyte used, which includes the electrostatic interaction of solvent/electrolyte with the iron centre and the cyclopentadienyl ring [14, 15].

Decamethylferrocene (DmFc or [FeII(η5-C5(CH3)5)2]) shows at least an order of magnitude weaker solvent- solute interaction compared to ferrocene, as the methyl- substituent groups in the cyclopentadienyl rings prevent both specific and non-specific interactions by hindering the access of organic solvent and supporting electrolyte molecules to the metal centre and to the cyclopentadienyl ring [14, 16, 17, 18, 19]. This observation was also confirmed after analysis of the X-ray diffraction patterns of decamethylferrocene, which indicate that the inter-ring methyl groups of DmFc are within Van der Waals distances [20]. Consequently, DmFc is better suited than Fc as an IRRS in organic solvents.

Decamethylferrocene is oxidized in organic solvents in a reversible, one-electron transfer, process to decamethylferrocenium (DmFc+ or [FeIII(η5-C5(CH3)5)2]+) and reduced back to DmFc according to Equation 1. The midpoint potential (Em), calculated from the average of

the oxidation (Epox) and reduction (Epred) peak potentials

[(Epox+Epred)/2],of this redox couple is more negative than

that of Fc as a consequence of the electron-donating effect

of the methyl groups (inductive effect), which by pushing

electron density towards the metal ion facilitates the

electron removal by the electrode. As given in Table 1, the

difference in the Em values between Fc0/+ and DmFc0/+

$$ v a r i e s f r o m a l o w o f 0. 4 1 3 \pm 0. 0 0 5 V i n $$

tetrahydrofuran/0.1M [Bu4N][BF4] to a high of 0.614 ±

0.005 V in dichloromethane/0.1M [Bu4N][TFAB] ([TFAB]

= [B(C6F5)4]–), which represents a difference of 0.201 V.

Furthermore, a variation of about 0.152 V can be

observed by changing the organic solvent from

tetrahydrofuran to 2,2,2-trifluoroethanol and keeping

constant the supporting electrolyte nature and concentration (0.1M [Bu4N][ClO4]). To minimize

uncertainties in the comparison of mid-point potentials in

Table 1, all potentials quoted are relative to the DmFc0/+

potential scale. Consequently, since the values presented

in this table represent the difference in Em, the variation

in the absolute values of Em of the Fc0/+ and DmFc0/+ IRRS

may probably be larger than those indicated here.

[FeII(η5-C5(CH3)5)2] Δ [FeIII(η5-C5(CH3)5)2]+ + e– (1)

SolventElectrolyteFc0/+ vs. DmFc0/+ (V)Ref
Diethyl ether0.1M [Bu N][BArF ] a
4 24		0.550±0.005[15]
0.1M Na[BArF ] a
24		0.583±0.005[15]
Anisole0.1M [Bu N][PF ]
4 6		0.518±0.005[15]
0.1M [Bu N][TFAB] b
4		0.607±0.005[15]
Methanol0.1M [Bu N][ClO ]
4 4		0.497±0.002[14]
Ethanol0.1M [Bu N][ClO ]
4 4		0.473±0.005[14]
2,2,2-trifluoroethanol0.1M [Bu N][ClO ]
4 4		0.575±0.004[14]
2-propanol0.1M [Bu N][CF SO ]
4 3 3		0.455±0.003[14]
Tetrahydrofuran0.1M [Bu N][BF ]
4 4		0.413±0.005[15]
0.1M [Bu N][CF SO ]
4 3 3		0.438±0.005[15]
0.1M [Bu N][ClO ]
4 4		0.423±0.005[15]
0.427±0.002[14]
0.1M [Bu N][PF ]
4 6		0.446±0.005[15]
0.1M [Bu N][BPh ]
4 4		0.485±0.005[15]
0.1M Na[BArF ] a
24		0.502±0.005[15]
0.1M [Bu N][TFAB] b
4		0.484±0.005[15]
0.1M [Bu N][BArF ] a
4 24		0.521±0.005[15]
Pyridine0.1M [Bu N][ClO ]
4 4		0.517±0.004[14]
0.1M [Bu N]Cl
4		0.534±0.005[15]
0.1M [Bu N][ClO ]
4 4		0.532±0.002[14]
0.1M [Bu N][PF ]
4 6		0.548±0.003[21]
0.1M [Et N][BF ]
4 4		0.541±0.003[22]
0.1M [Bu N][TFAB] b
4		0.614±0.005[15]
0.1M [C mPyr][FAP]
4		0.589±0.003[22]

Table 1: Redox potentials of ferrocene in different organic media and supporting electrolytes.

Medicinal & Analytical Chemistry International Journal

Dichloromethane0.1M [C mim][FAP]
2		0.590±0.003[22]
0.1M [C mim][B(CN) ]
2 4		0.588±0.003[22]
0.1M [C mim][N(Tf) ]
4 2		0.570±0.003[22]
0.1M [C mPyr][N(Tf) ]
4 2		0.568±0.003[22]
0.1M [C mim][FSI]
2		0.569±0.003[22]
0.1M [C mim][FSI]
3		0.568±0.003[22]
0.1M [C mPyr][N(CN) ]
4 2		0.564±0.003[22]
0.1M [C mim][PF ]
4 6		0.556±0.003[22]
0.1M [C mim][BF ]
4 4		0.557±0.003[22]
0.1M [C mim][CF SO ]
4 3 3		0.556±0.003[22]
1,2-dichloroethane0.1M [Bu N][ClO ]
4 4		0.532±0.001[14]
1,2-dibromoethane0.1M [Bu N][ClO ]
4 4		0.475±0.007[14]
Benzonitrile0.1M [Bu N]Cl
4		0.524±0.005[15]
0.1M [Bu N][ClO ]
4 4		0.523±0.001[14]
0.1M [Bu N][PF ]
4 6		0.530±0.005[15]
0.1M [Bu N][TFAB] b
4		0.543±0.005[15]
Bromobenzene0.1M [Bu N][ClO ]
4 4		0.489±0.005[14]
Chlorobenzene0.1M [Bu N][ClO ]
4 4		0.497±0.001[14]
1,2-dichlorobenzene0.1M [Bu N][ClO ]
4 4		0.535±0.001[14]
Benzyl alcohol0.1M [Bu N][ClO ]
4 4		0.508±0.003[14]
Nitrobenzene0.1M [Bu N][ClO ]
4 4		0.514±0.002[14]
Aniline0.1M [Bu N][ClO ]
4 4		0.527±0.004[14]
Toluene[Bu N][BF ] c
4 4		0.430±0.005[15]
Acetone0.1M [Bu N]Cl
4		0.451±0.005[15]
0.1M [Bu N][ClO ]
4 4		0.479±0.004[14]
0.1M [Bu N][PF ]
4 6		0.487±0.005[15]
0.1M [Bu N][TFAB] b
4		0.504±0.005[15]
Propylene carbonate0.1M [Bu N][ClO ]
4 4		0.495±0.002[14]
Acetonitrile0.1M [Bu N]Cl
4		0.501±0.005[15]
0.1M [Bu N][ClO ]
4 4		0.505±0.002[14]
0.1M [Bu N][PF ]
4 6		0.509±0.003[21]
0.1M [Bu N][TFAB] b
4		0.517±0.005[15]
Nitromethane0.1M [Bu N]Cl
4		0.505±0.005[15]
0.1M [Bu N][ClO ]
4 4		0.516±0.004[14]
0.1M [Bu N][PF ]
4 6		0.510±0.005[15]
0.1M [Bu N][TFAB] b
4		0.516±0.005[15]
Formamide0.1M [Bu N][ClO ]
4 4		0.510±0.003[14]
N-methylformamide0.1M [Bu N][ClO ]
4 4		0.510±0.002[14]
N,N-dimethylformamide0.1M [Bu N]Cl
4		0.475±0.005[15]
0.1M [Bu N][ClO ]
4 4		0.458±0.003[14]
0.1M [Bu N][PF ]
4 6		0.478±0.005[15]
0.1M [Bu N][TFAB] b
4		0.493±0.005[15]
N,N-dimethylacetamide0.1M [Bu N][ClO ]
4 4		0.455±0.008[14]
Dimethyl sulfoxide0.1M [Bu N][PF ]
4 6		0.486±0.005[15]
0.1M [Bu N][TFAB] b
4		0.493±0.005[15]
0.1M [Bu N][ClO ]
4 4		0.468±0.001[14]
Chloroform0.1M [Bu N][ClO ]
4 4		0.483±0.001[14]

Table 2: Redox potentials of ferrocene in different organic media and supporting electrolytes.

a[BArF24] = [B(C6H3(CF3)2)4]–; b [TFAB] = [B(C6F5)4]–; c the toluene:[Bu4N][BF4] electrolyte is of the 3:1 stoichiometry. Table 1: Redox potentials of ferrocene in different organic media and supporting electrolytes.

Medicinal & Analytical Chemistry International Journal

The solvent-Fc interaction is also present in ionic liquids. This issue was addressed by studying the effect of ionic liquid structure on the mid-point potentials of Fc and DmFc in eleven different ILs as well as in dichloromethane with added IL as the supporting electrolyte [22]. As evidenced in Table 2, a variation in the Em of Fc vs. DmFc0/+ of about 0.100 V is observed when the IL was changed from [C4mim][CF3SO3] to [C2mim][FAP] under neat conditions (water content < 100 ppm). Similarly, when DmFc and Fc were simultaneously present in dichloromethane with added IL as the supporting electrolyte-conditions under which DmFc is less sensitive to solvation effects-a variation in the Em of Fc vs. DmFc0/+ of about 0.050 V was observed when the supporting electrolyte was changed from 0.1 M [Et4N][BF4] to 0.1 M [C2mim][FAP]. A variable potential difference which increases from 0.014 V in the case of [C4mPyr][FAP] to 0.082 V in the case of [C4mim][CF3SO3] is noticed from the comparison of the Em of Fc vs. DmFc0/+ between neat ILs and diluted conditions (dichloromethane with added IL as the supporting electrolyte) [22].

ILE vs. DmFc0/+ (V)
m		E vs. Fc0/+ (V)
m		Ref
FcCc+Cc+
[C mPyr][FAP]
4		0.575 a-0.747 a-1.322 a[1,22]
[C mim][FAP]
2		0.574 a-0.749 a-1.323 a[1,22]
[C mim][B(CN) ]
2 4		0.526 a-0.799 a-1.325 a[1,22]
[C mim][N(Tf) ]
2 2		0.520 a-0.820 a-1.330 a[23,24]
[C mim][N(Tf) ]
4 2		0.514 a-0.813 a-1.327 a[1,22]
[C mPyr][N(Tf) ]
3 2		-1.325 b[25]
-1.327 b[25]
[C mPyr][N(Tf) ]
4 2		0.511 a-0.816 a-1.327 a[1,22]
-1.363 b[25]
-1.333 d[26]
[S ][N(Tf) ]
2 2 1 2		-1.348 b[25]
[S ][N(Tf) ]
2 2 2 2		-1.354 b[25]
[C mpip][N(Tf) ]
3 2		-1.362 b[25]
[HmimSC ][N(Tf) ]
4 2		0.510 a-0.830 a-1.330 a[23]
[C mim][FSI]
2		0.512 a-0.820 a-1.332 a[1,22]
[C mPyr][FSI]
3		0.510 a-0.822 a-1.332 a[1,22]
[C mim][N(CN) ]
2 2		-1.332 b[25]
-1.355 b[25]
[C mPyr][N(CN) ]
4 2		0.505 a[1,22]
[C mim][PF ]
4 6		0.478 a-0.851 a-1.329 a[1,22]
[C mim][BF ]
2 4		-1.336 c[27]
[C mim][BF ]
4 4		0.478 a-0.848 a-1.326 a[1,22]
[C mim][CF SO ]
4 3 3		0.474 a-0.854 a-1.328 a[1,22]
-1.362 b[25]
[C mPyr][CF SO ]
4 3 3		-1.347 b[25]

Table 3: Redox potentials of transition-metal sandwich complexes obtained by cyclic voltammetry in different ionic liquids.

aPotentials were obtained using a scan rate of 0.1 Vs-1. Em values were obtained with an accuracy of ± 0.003 V (n=5). bNo experimental error was reported. c_E_m values were obtained with an accuracy of ± 0.002 V. d_E_m value was obtained with an accuracy of ± 0.001 V. Table 2: Redox potentials of transition-metal sandwich complexes obtained by cyclic voltammetry in different ionic liquids.

The cobaltocenium|cobaltocene, Cc+/0 (also known as [CoIII(η5-C5H5)2]+/[CoII(η5-C5H5)2]), redox couple was used in organic solvents as an alternative IRRS. It was suggested by Strehlok, et al. in 1960 as an IRRS for organic solvent systems [28]. Although it was not considered by IUPAC as an alternative to the Fc0/+ couple, it shows negligible volatility and high stability and solubility in some organic solvents [29] and ILs [12, 23, 30, 31]. If the cathodic potential window of the solvent is large enough, two chemically reversible, one- Medicinal & Analytical Chemistry International Journal

electron reduction processes can be observed (Eqs. 2 and 3).

[CoIII(η5-C5H5)2]+ + e- Δ [CoII(η5-C5H5)2] (2) [CoII(η5-C5H5)2] + e- Δ [CoI(η5-C5H5)2]– (3) The midpoint potential for the Cc0/- couple (Eq 3) was found to be considerably solvent dependent and not always can be determined accurately at room temperature, mainly due to the overlap and interaction of this process with the electrochemical response of the organic solvent used [29]. Thereby, only the Cc+/0 redox couple (Eq 2) has been extensively used in organic solvents to provide a known and stable reference point. As an extrapolation of concept, it is also used in IL systems. Surprisingly, little is known about the solvation effect of either different organic solvent with added supporting electrolytes or IL structure on the Cc+/0 formal potential [2].

Table 2 reports the IL effect on the Cc+/0 process using both Fc0/+ and DmFc0/+ as the IRRS. In the first case, the Em of Cc+/0 is almost constant and could be represented by a value of 1.327 ± 0.005 V vs. Fc0/+ in 14 different ILs and using glassy carbon as the working electrode (exceptions are [C2mim][BF4] and [C4mPyr][CF3SO3], where a value of 1.336 and 1.347 V was reported, respectively). Meanwhile, when DmFc0/+ is used as the IRRS, an increase of 0.107 V in the potential separation between Cc+/0 and DmFc0/+ couples is observed upon changing the IL from [C4mPyr][FAP] to [C4mim][CF3SO3].

To further understand this observation, we need to compare the variation in the Em of Cc+/0 vs. Fc0/+ in organic solvents with added supporting electrolyte. It is well known that Fc is highly sensitive to solvation effects under these conditions (see above). As possible to see in Table 3, the Em of Cc+/0 in 9 different solvent/supporting electrolyte combinations can be represented by an almost constant value of about 1.352 ± 0.015 V vs. Fc0/+. As saw in Table 1, the Em value of the Fc0/+ redox process varies from 0.437 ± 0.005 V in ethanol/0.1M [Bu4N][ClO4] to a value of 0.532 ± 0.002 V in dichloromethane/0.1M [Bu4N][ClO4], which represents a difference of 0.095 V. Meanwhile, the obtained Cc+ Em values vs. Fc0/+ in the same two organic solvent systems (with 0.1M [Bu4N][ClO4] as the supporting electrolyte) show a variation of just 0.008 V (Table 3).

SolventElectrolyteCc+/0 vs. Fc0/+ (V)Ref
Acetonitrile0.1M [Et N][ClO ]
4 4		-1.350±0.003[29]
0.1M [Bu N][BF ]
4 4		-1.348±0.003[29]
0.1M [Bu N][PF ]
4 6		-1.337±0.005
Dichloromethane0.1M [Et N][ClO ]
4 4		-1.359±0.007[29]
0.1M [Bu N][BF ]
4 4		-1.355±0.007[29]
ethanol0.1M [Et N][ClO ]
4 4		-1.351±0.004[29]
0.1M [Bu N][BF ]
4 4		-1.356±0.004[29]
Water0.1M Li[ClO ]
4		-1.367±0.002[29]
Toluene0.4M [P ][FAP]
6 6 6 14		-1.340±0.005[32]

Table 4: Redox potentials of cobaltocenium hexafluorophosphate in different organic media and supporting electrolytes.

[P6 6 6 14][FAP] = trihexyl(tetradecyl)phosphonium tris(pentafluoroethyl)trifluorophosphate. Table 3: Redox potentials of cobaltocenium hexafluorophosphate in different organic media and supporting electrolytes.

Other alternatives, such as 1,1’-dimethylferrocene, 1,1- diacetylferrocene, ferrocenecarboxaldehyde, etc., could be

Limitations in the Application of IRRS in Ionic Liquids

  • Poor solubility of Fc and DmFc in some ionic liquids, mostly in those with high viscosity [31, 34, 35].
  • Moderate volatility of Fc [35, 36, 37].
  • Reactivity of Fc+ with the IL components: generally observed when the anodic potential window of the IL is close to the formal redox potential of Fc0/+ [23, 31].

Medicinal & Analytical Chemistry International Journal

  • The Em of Fc0/+ redox process differs from the Eo′: this effect is related to the inequality in the diffusion coefficients of Fc and Fc+ in almost all known ILs [11, 35].
  • Reactivity of decamethylferrocene in the presence of oxygen [34].

Selection of an Internal Reference Redox Scales for Voltammetric Measurements

As a conclusion, good practice for electrochemistry dictates that the appropriated selection of the IRRS to provide a known and stable reference point in voltammetry is very important both for organic and IL solvent systems. The objective is to choose the IRRS (either the OXIRRS or REDIRRS component of a reversible redox couple) so that the cyclic voltammetric response for the simultaneously present electroactive ANS can be observed independently of the IRRS response. This means that if the ANS to be studied is in the oxidized form, then the reduced form of the IRRS must be added to the system. Conversely, if the ANS is in the reduced form, then the oxidized form of the IRRS must be added to the system. However, it is possible to see a large number of published works, which do not follow this basic rule [10].

The redox potential of Fc and Cc+ are sensitive to the solvation properties of the organic solvent or IL components. Consequently, the use of Fc+/0 and Cc+/0 couples as IRRS is expected to be adequate only when a single solvent system is being studied. However, corrections may be required in conditions that require comparison between different solvent systems. Even though DmFc is not freely soluble in ILs, the existing experimental results suggest that the DmFc0/+ redox process is a suitable IRRS for voltammetric studies both in organic solvents and in ionic liquids, as it is less dependent on the solvent and supporting electrolyte nature in comparison to what is observed for ferrocene and cobaltocenium.

References

  1. As a conclusion, good practice for electrochemistry dictates that the appropriated selection of the IRRS to provide a known and stable reference point in voltammetry is very important both for organic and IL solvent systems. The objective is to choose the IRRS (either the OXIRRS or REDIRRS component of a reversible redox couple) so that the cyclic voltammetric response for the simultaneously present electroactive ANS can be observed independently of the IRRS response. This means that if the ANS to be studied is in the oxidized form, then the reduced form of the IRRS must be added to the system. Conversely, if the ANS is in the reduced form, then the oxidized form of the IRRS must be added to the system. However, it is possible to see a large number of published works, which do not follow this basic rule [10]. The redox potential of Fc and Cc+ are sensitive to the solvation properties of the organic solvent or IL components. Consequently, the use of Fc+/0 and Cc+/0 couples as IRRS is expected to be adequate only when a single solvent system is being studied. However, corrections may be required in conditions that require comparison between different solvent systems. Even though DmFc is not freely soluble in ILs, the existing experimental results suggest that the DmFc0/+ redox process is a suitable IRRS for voltammetric studies both in organic solvents and in ionic liquids, as it is less dependent on the solvent and supporting electrolyte nature in comparison to what is observed for ferrocene and cobaltocenium.
More from this journal

Cite this article

BibTeX
APA
RIS
@article{angel2019,
  title   = {On Choosing Ferrocene as an Internal Reference Redox Scale for Voltammetric Measurements: A Cautionary Tale},
  author  = {Angel AJ Torriero},
  journal = {Medicinal & Analytical Chemistry International Journal},
  year    = {2019},
  volume  = {3},
  number  = {4},
  doi     = {10.23880/macij-16000151}
}
Angel AJ Torriero (2019). On Choosing Ferrocene as an Internal Reference Redox Scale for Voltammetric Measurements: A Cautionary Tale. Medicinal & Analytical Chemistry International Journal, 3(4). https://doi.org/10.23880/macij-16000151
TY  - JOUR
TI  - On Choosing Ferrocene as an Internal Reference Redox Scale for Voltammetric Measurements: A Cautionary Tale
AU  - Angel AJ Torriero
JO  - Medicinal & Analytical Chemistry International Journal
PY  - 2019
VL  - 3
IS  - 4
DO  - 10.23880/macij-16000151
ER  -