Fibre optic ATR-IR spectroscopy at cryogenic temperatures: In-line reaction monitoring on organolithium compounds

Article abstract of DOI:10.1039/c2cc16016a

A reliable methodology, utilising an ATR-IR fibre probe, for in-line monitoring of low temperature reactions is presented. The application of this convenient set-up enables a fast and safe exploration of highly reactive substrates. Hence, in situ monitoring of lithiation reactions is realised and the potential to investigate sensitive intermediates is being demonstrated. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2cc16016a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 2451–2453
www.rsc.org/chemcomm
COMMUNICATION
Fibre optic ATR-IR spectroscopy at cryogenic temperatures: in-line
reaction monitoring on organolithium compoundsw
a
b
a
a
b
¨
Daniel Lumpi,* Christoph Wagner, Matthias Schopf, Ernst Horkel, Georg Ramer,
a
b
¨
Bernhard Lendl and Johannes Frohlich*
Received 27th September 2011, Accepted 6th January 2012
DOI: 10.1039/c2cc16016a
A reliable methodology, utilising an ATR-IR fibre probe, for in-line
monitoring of low temperature reactions is presented. The application
of this convenient set-up enables a fast and safe exploration of highly
reactive substrates. Hence, in situ monitoring of lithiation reactions is
realised and the potential to investigate sensitive intermediates is
being demonstrated.
In this contribution a reliable methodology in order to
realise probe applications at cryogenic temperatures is being
developed. The demonstrated investigations on organolithium
intermediates outline their potential impact on routine synthetic
steps and also to derive kinetic data of highly complex multi-step
reactions suitable for mechanistic considerations.
A key challenge in spectroscopic measurements in reaction
mixtures at cryogenic temperatures is that dynamic processes can
easily lead to the sample spectra being acquired at substantially
different temperatures than the background. Applying the
aforementioned mid-IR fibre optic probe exothermic effects,
and also simple reactant addition in the course of the synthesis,
resulted in spectra heavily overlaid with (co-)sine-type artifacts,
rendering the acquired spectra useless for detailed interpretation.
These artifacts are attributed to differing thermal expansion
coefficients of the probe material and the diamond ATR
crystal and probably represent a common problem preventing
the application of IR monitoring in fast dynamic processes at
low temperature. Various standard correction approaches7 were
applied to reduce these fringes, but none yielded satisfactory results.
Thus, a dedicated procedure based on Fast Fourier Transformation
(FFT) was developed and implemented to the data evaluation
process. In this approach the correction was realised by applying
an FFT on the recorded spectra, which deconvolutes the data into a
linear combination of sine and cosine functions of different
frequencies. To remove the introduced artifacts the Fourier
coefficients of the corresponding low frequencies were set to zero.
After an application of this high pass filter and retransformation by
inverse FFT, the fringes in the spectra were significantly reduced.
The number of removed Fourier coefficients can be adjusted
individually, depending on the encountered temperature differences
for each spectrum specifically. Errors in absorption and frequency
values observed during the FFT correction as well as further
information and limitations of this strategy are given in the ESI.w
Applying the FFT filter rendered it possible to acquire reliable IR
spectroscopic data of organolithium compounds during reaction
progress at temperatures even below ꢀ80 1C covering a spectral
range from 600 to 1800 wavenumbers.
Reaction conditions in synthesis are commonly optimised
using off-line methods to determine operation parameters.1
However, on-line or especially in-line methods such as time
resolved spectroscopy can provide highly relevant information
for process optimisation and scale-up.1 Still, dynamic reaction
analysis in order to monitor highly reactive species at low
temperatures (e.g. lithiation chemistry) remains a challenging task.
Organolithium compounds gained outstanding importance as
powerful reagents and key intermediates in organic synthetic
chemistry.2 Therefore, since the discovery by Schlenk and Holtz,3
lithiation chemistry has grown into a well established technique
also receiving attention in modern industrial synthesis.4 As a
consequence of the highly reactive intermediates such lithiation
reactions are generally performed at low temperatures of
approximately ꢀ60 to ꢀ80 1C, in some cases even below ꢀ100 1C.2
Encouraged by recent studies,5 a mid-IR fibre optic probe
comprising an attenuated total reflection (ATR) element was
selected for a fast and safe in-line monitoring of lithiation
reactions. The major advantage of this in situ technique versus
off-line approaches is that monitoring takes place inside the
reaction system, thus eliminating risks of sample alteration
during probing. This method proved to be a versatile and effective
tool to explore intermediates at elevated temperatures,6 however,
to the best of our knowledge, no examples of applications
towards reactive metal halogen (inter-)exchange reactions at
cryogenic temperatures have been reported up to date.
a Institute of Applied Synthetic Chemistry, Vienna University of
Technology, Getreidemarkt 9/163-OC, 1060 Wien, Austria.
E-mail: daniel.lumpi@tuwien.ac.at, johannes.froehlich@tuwien.ac.at;
Fax: +43 1 58801 154 99; Tel: +43 1 58801 163 719
b Institute of Chemical Technologies and Analytics, Vienna University
of Technology, Getreidemarkt 9/164-AC, 1060 Wien, Austria
w Electronic supplementary information (ESI) available: Experimental
details, additional information regarding the FFT correction procedure
and the double-sided halogen dance reaction as well as spectral data.
See DOI: 10.1039/c2cc16016a
To follow the formation of organolithium species, integrated
values (areas) under strong non-interfered absorption bands
were evaluated subsequently to the FFT correction.8 As a first
step the metal halogen exchange reaction on aryl halides was
investigated as mentioned above.8 The results show very high
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 2451–2453 2451

View Article Online
reaction rates in THF at ꢀ86 1C reaching a steady state in the
phenyllithium concentration within 1.0 min for bromobenzene,
and within 30 s when using iodobenzene.8 In comparison,
protocols published in the recent literature report reaction times
of 30–60 min for bromobenzene and nBuLi in THF at ꢀ78 1C.9
This clearly illustrates the potential to effectively optimise
reaction times also in relatively fast processes under cryogenic
conditions. Metal halogen exchange reactions investigated on
bromothiophene derivatives are presented in the ESI.w
Scheme 1 Investigation on activating effects; metallation reactions
using LDA on (bromo-)thiophene derivatives.
Slower lithiations were observed using Et2O as the solvent,
yielding no reaction of bromobenzene and a significantly lower
reaction rate for iodobenzene at ꢀ86 1C (Fig. 1). Consequently, an
increase in the reaction temperature also yielded good results for
the FFT method. In Et2O at ꢀ40 1C bromobenzene is consumed
relatively slowly; in contrast the reaction of iodobenzene is
accomplished in less than 1.0 min. Reasonable conversion
rates of bromobenzene in Et2O could be observed at 2 1C
(state of equilibrium: B1400 s). These results correspond to data
previously reported at various temperatures or by applying other
techniques (e.g. rapid-injection NMR10).2 The significant tempera-
ture dependence of reaction rates clearly points out the importance
of in-line monitoring, as both sampling and standard bypass
approaches cannot ensure constant thermal conditions in the course
of analysis, thus leading to erroneous readings—especially when
directly compared to the actual (batch) process. In contrast the
application of IR probe spectroscopy allows monitoring of highly
sensitive reactants both in situ and in real-time.
The Halogen Dance (HD) reaction11 represents a powerful
tool, especially in heterocyclic chemistry, to reliably achieve
specific substitution patterns. In a preliminary study we intended
to clarify the influence introduced by the migrating halogen on
the lithiation. Therefore, the activating effects of the bromine
group on the metallation reaction of thiophene moieties were
studied for a lithium diisopropylamide (LDA) induced lithiation
of 2- and 3-bromothiophene, respectively (Scheme 1).
Fig. 2 Effects of activating halide groups on thiophene moieties in
metallation reactions using LDA at ꢀ86 1C and ꢀ40 1C (solvent: THF).
double- vs. single-sided HD reactions (Scheme 2). In this
multi-step sequence the first metallation and the HD are
expected to proceed faster, leading to accumulation of 2 before
the second lithiation towards 3 occurs.12 However, kinetic
data of the in situ formation of 2 during a double-sided
Halogen Dance are of great value for synthetic aspects and
possible application of this reaction.
By in-line ATR-IR monitoring, again applying FFT correction,
the accumulation of 2 can be clearly observed (Fig. 3A). With the
results indicating a very good agreement of the time constants
(31.8 s for LDA vs. 31.1 s for 2), the comparison of the kinetic
parameters of LDA consumption and intermediate formation
confirms the formation of 2 as being the first step in the sequence
(Fig. 3B). This proves that in the first reaction step LDA is
exclusively consumed to form 2; the second HD is realised in a
next step leading to 3. Complementarily,8 we investigated the ratio
of single- and double-sided HD (2 and 3, respectively) depending
on the amount of LDA via NMR spectroscopy, supporting the
assumption (Scheme 2) of the metallation being the rate
determining step in this HD reaction; the migration of the
halogen proceeds relatively fast (no accumulation of
3-lithiothiophene species, for details see ESIw).
As obvious from Fig. 2 a strong activating effect of the
bromine was determined. In THF at ꢀ86 1C no reaction
progress could be observed on thiophene; in comparison using
2- and 3-bromothiophene good conversion rates were detected.
This highly increased reactivity exerted by inductive effects of
the halogen substituent is surprising especially in the case of
2-bromothiophene; the faster conversion of 3-bromo- compared to
2-bromothiophene is attributed to a stronger inductive effect of the
bromine substituent in the ortho- compared to the para-position.
Metallation of thiophene in THF was observed at ꢀ40 1C.
In previous work we examined halogen migrations on
5,50-dibromo-2,20-bithiophene focusing on the selectivity of
Accessing kinetic parameters for the formation of intermediate
3 required multivariate data analysis due to overlapping of
absorption bands of 3 (B1017 cmꢀ1) with an LDA absorbance
Scheme 2 Double-sided Halogen Dance reaction on 5,50-dibromo-
2,20-bithiophene utilising LDA (2.5 equiv.) at ꢀ40 1C. (i) Metallation
reaction, (ii) Halogen Dance reaction, (iii) excess of methanol.
Fig. 1 Metal halogen exchange reaction of bromo- and iodobenzene
at different temperatures (nBuLi, ꢀ86 1C and ꢀ40 1C, Et2O).
c
2452 Chem. Commun., 2012, 48, 2451–2453
This journal is The Royal Society of Chemistry 2012

View Article Online
the possibility to disclose kinetic data, also of sensitive reaction
intermediates, was demonstrated to be an important factor for
e.g. mechanistic considerations or process optimisation.
In conclusion not only information about lithiation reactions is
presented but also a methodology, based on an FFT procedure, for
an effective reaction monitoring at cryogenic temperature widely
applicable in the field of chemistry is developed. This is in particular
true since mid-IR fibre optic probes are nowadays commercially
available.14 Potential applications in the field of organic synthetic
chemistry are widespread; for example aromatic substitution,
oxidation and selective hydrogenation as well as reactions of
sensitive substrates in general are of great interest concerning
investigations at low temperatures. Furthermore, the compatibility
to heterogeneous reaction mixtures, high pressure conditions and
biotechnological applications expands the scope of this technique.
The authors thank A.R.T. Photonics for providing
the applied mid-IR fibre optic probe; M. Brandstetter and
M. Kraft for fruitful discussion. Financial support from
Carinthian Tech Research AG and the COMET program of
the Austrian government is also gratefully acknowledged.
Notes and references
Fig. 3 In-line IR sensor spectra during the double-sided Halogen Dance
reaction (ꢀ40 1C, t = 0 s: 1 already included, t = 12 s: LDA addition);
(A) 3D plot of spectra (750–1050 cmꢀ1) recorded during reaction
progress (0–160 s), (B) monitoring of Li-species LDA and intermediate
2; exp. fit LDA (t1: 31.8 s, R2: 0.99), exp. fit 2 (t1: 31.1, R2: 0.99).
1 A. E. Rubin, S. Tummala, D. A. Both, C. Wang and E. J. Delany,
Chem. Rev., 2006, 106, 2794; and references therein.
2 (a) B. J. Wakefield, The Chemistry of Organolithium Compounds,
Pergamon, Oxford, 1974; (b) M. Schlosser, in Organometallics in
Synthesis: A Manual, ed. M. Schlosser, Wiley, Chichester, 2002,
p. 1; (c) J. Clayden, Organolithiums: Selectivity for Synthesis,
Elsevier Pergamon, Oxford, 2002; (d) Z. Rappoport and
I. Marek, The Chemistry of Organolithium Compounds, Wiley,
Chichester, 2004; and references therein.
3 W. Schlenk and J. Holtz, Ber. Dtsch. Chem. Ges., 1917, 50, 262.
4 (a) G. Wu and M. Huang, Chem. Rev., 2006, 106, 2596, and
references therein; (b) T. L. Rathman and W. F. Bailey, Org.
Process Res. Dev., 2009, 13, 144.
5 (a) H. M. Heise, L. Kuepper and L. N. Butvina, Anal. Bioanal.
Chem., 2003, 375, 1116; (b) E. Weymeels, H. Awad, L. Bischoff,
F. Mongin, F. Trecourt, G. Queguiner and F. Marsais, Tetrahedron,
2005, 61(13), 3245; (c) U. Bentrup, L. Kuepper, U. Budde, K. Lovis
and K. Jaehnisch, Chem. Eng. Technol., 2006, 29, 1216;
(d) C. B. Minnich, P. Buskens, H. C. Steffens, P. S. Baeuerlein,
L. N. Butvina, L. Kuepper, W. Leitner, M. A. Liauw and L. Greiner,
Org. Process Res. Dev., 2007, 11, 94; (e) L. Gupta, A. C. Hoepker,
K. J. Singh and D. B. Collum, J. Org. Chem., 2009, 74(5), 2231.
6 D. Lumpi, C. Braunshier, C. Hametner, E. Horkel, B. Zachhuber,
at B1008 cmꢀ1. Hence, multivariate curve resolution using
alternating least squares (MCR-ALS)13 algorithm was applied
to the spectral data set (Fig. 3A). This method decomposes the
data into species spectra and their corresponding concentration
profiles. Fig. 4 shows the result of the MCR-ALS analysis for
intermediate species 2 and 3 including bi-exponential fits to the
data points. The results again outline the good match of kinetic
parameter between the consumption of 2 and the formation of
3, approving the aforementioned conclusions. Moreover, the
good agreement of data in the first reaction step between
integrated areas (Fig. 3B) and the MCR-ALS algorithm
(Fig. 4) underlines the applicability of this multivariate tool.
Summarising, a reliable methodology enabling successful
monitoring of reactions under cryogenic conditions in real-time is
presented. It was shown that this in-line technique utilising an
FFT correction procedure is suitable for a convenient and safe
exploration of (hetero-)aromatic lithium compounds. In addition,
B. Lendl and J. Frohlich, Tetrahedron Lett., 2009, 50, 6469.
¨
7 Utilisation of correction methods integrated in the OPUS 6.5
(Bruker Optik GmbH, Ettlingen, Germany) software package.
8 Further information is provided in the ESIw.
9 (a) M. P. Jensen, S. J. Lange, M. P. Mehn, E. L. Que and L. Que,
Jr., J. Am. Chem. Soc., 2003, 125, 2113; (b) S.-J. Liu, Q. Zhao,
Q.-L. Fan and W. Huang, Eur. J. Inorg. Chem., 2008, 2177;
(c) Y.-I. Park, J.-H. Son, J.-S. Kang, S.-K. Kim, J.-H. Lee and
J.-W. Park, Chem. Commun., 2008, 2143.
10 J. F. McGarrity, J. Prodolliet and T. Smith, Org. Magn. Reson.,
1981, 17, 59.
11 (a) M. Schnuerch, M. Spina, A. Farooq Khan, M. D. Mihovilovic
and P. Stanetty, Chem. Soc. Rev., 2007, 36, 1046; (b) M. Vinicius
Nora de Souza, Curr. Org. Chem., 2007, 11(7), 637; (c) J. Froehlich,
Prog. Heterocycl. Chem., 1994, 6, 1; (d) M. Schlosser, Angew. Chem.,
Int. Ed., 2005, 44, 376.
12 R. Bobrovsky, C. Hametner, W. Kalt and J. Froehlich, Heterocycles,
2008, 76(2), 1249.
13 J. Jaumot, R. Gargallo, A. de Juan and R. Tauler, Chemom. Intell.
Lab.Syst., 2005, 76, 101.
14 Mettler-Toledo Int. Inc. (Greifensee, Switzerland), A.R.T. Photonics
GmbH (Berlin, Germany), IFS Aachen (Aachen, Germany).
Fig. 4 Intermediate 3 formation as extracted from the spectral data
set via MCR-ALS algorithm; bi-exp. fit 2 (t1: 6.9 s, t2: 2.5 ꢁ 10ꢀ6 s,
R2: 0.98), bi-exp. fit 3 (t1: 6.8 s, t2: 3.3 ꢁ 10ꢀ6 s, R2: 0.98).
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 2451–2453 2453