A new approach for the detection of ethylene using silica-supported palladium complexes

Article abstract of DOI:10.1021/ic800986t

The coordination of olefins to square-planar Pd(II) and Pt(II) complexes containing 2,9-dimethylphenanthroline (L1) often involves a change of color associated with a change of geometry at the metal center. In order to obtain suitable colorimetric detectors for ethylene gas, a series of new Pd(II) and R(II) compounds with a range of 2,9-disubstituted phenanthroline ligands [2,9-di-n-butyl-1,10-phenanthroline (L2), 2,9-di-s-butyl-1,10-phenanthroline (L3), 2,9-diphenyl-1,10-phenanthroline (L4), and 2,9-dimethyl-4,7-diphenyl-1,10- phenanthroline (bathocuproine, L5)] have been prepared and their reactivity toward ethylene investigated both in solution and after depositing the detector compounds on a variety of solid supports. The Pd(II) complex [PdCl ₂(L2)] supported on silica undergoes a clear color change upon exposure to ethylene, while remaining stable toward air and water, and forms the basis for new simple colorimetric detectors with potential applications in ethylene pipe-leak detection and the monitoring of fruit ripening. Encouragingly, the detector is able to discriminate between fruit at different stages of ripening. The response of the detector to other volatiles was also examined, and specific color changes were also observed upon exposure to aromatic acetylenes. The crystal structures of four new derivatives, including the ethylene-Pt(II) complex [PtCl₂(C₂H₄)(L2)], are also described.

Full text of DOI:10.1021/ic800986t

Inorg. Chem. 2008, 47, 9035-9041
A New Approach for the Detection of Ethylene Using Silica-Supported
Palladium Complexes
Patricia Cabanillas-Galán,† Linda Farmer,‡ Terence Hagan,‡ Mark Nieuwenhuyzen,†
Stuart L. James,*,† and M. Cristina Lagunas*,†
School of Chemistry and Chemical Engineering, Queen’s UniVersity Belfast, Stranmillis Road,
Belfast BT9 5AG, United Kingdom, and Food & EnVironmental Science DiVision, Newforge Lane,
Belfast BT9 5PX, United Kingdom
Received June 6, 2008
The coordination of olefins to square-planar Pd(II) and Pt(II) complexes containing 2,9-dimethylphenanthroline (L1)
often involves a change of color associated with a change of geometry at the metal center. In order to obtain
suitable colorimetric detectors for ethylene gas, a series of new Pd(II) and Pt(II) compounds with a range of 2,9-
disubstituted phenanthroline ligands [2,9-di-n-butyl-1,10-phenanthroline (L2), 2,9-di-s-butyl-1,10-phenanthroline (L3),
2,9-diphenyl-1,10-phenanthroline (L4), and 2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline (bathocuproine, L5)] have
been prepared and their reactivity toward ethylene investigated both in solution and after depositing the detector
compounds on a variety of solid supports. The Pd(II) complex [PdCl2(L2)] supported on silica undergoes a clear
color change upon exposure to ethylene, while remaining stable toward air and water, and forms the basis for new
simple colorimetric detectors with potential applications in ethylene pipe-leak detection and the monitoring of fruit
ripening. Encouragingly, the detector is able to discriminate between fruit at different stages of ripening. The response
of the detector to other volatiles was also examined, and specific color changes were also observed upon exposure
to aromatic acetylenes. The crystal structures of four new derivatives, including the ethylene-Pt(II) complex
[PtCl2(C2H4)(L2)], are also described.
Introduction
mable, and can form explosive mixtures in the air (explosive
limits: 2.7-36 vol % in the air). The laser-based gas detectors
currently used industrially are relatively expensive and
complicated to use, with monitoring typically taking place
once a week. Inexpensive colorimetric leak detectors will
therefore have significant safety and economic implications
for the industries that produce or use the gas. In addition,
ethylene is emitted by climacteric fruit (such as apples,
bananas, or tomatoes) as part of the ripening process,3,4 and
visual detectors for ethylene are therefore also of interest as
indicators of food ripeness (for example, in “smart packag-
ing”). Permanganate5 and Pd(II)/molybdate6 have been put
forward as sensor materials for ethylene in the context of
fruit ripening. These methods rely on the color-changing
There is a need for inexpensive, fast, simple, and sensitive
monitoring devices for the detection of volatile compounds,
for example, for the environmental and occupational moni-
toring of emissions. Optical monitoring based on color
changes is one of the simplest approaches to gas monitoring
and can be very low-cost and highly analyte-specific.1 In
particular, a visual detector for ethylene would be of interest
because it is the largest-volume organic chemical produced
worldwide (>70 million tonnes in 20062), is highly flam-
*Authorstowhomcorrespondenceshouldbeaddressed.E-mail:c.lagunas@
qub.ac.uk (M.C.L.), s.james@qub.ac.uk (S.L.J.).
†
Queen’s University Belfast.
Food & Environmental Science Division.
‡
reduction of MnO4 or Mo6+, but neither of these systems
-
(1) See, for example: (a) Martinez, A. W.; Phillips, S. T.; Butte, M. J.;
Whitesides, G. M. Angew. Chem., Int. Ed. 2007, 46, 1318–1320. (b)
Zhang, C.; Suslick, K. S. J. Agric. Food Chem. 2007, 55, 237–242.
(c) Janzen, M. C.; Ponder, J. B.; Bailey, D. P.; Ingison, C. K.; Suslick,
K. S. Anal. Chem. 2006, 78, 3591–3600. (d) Zhang, C.; Suslick, K. S.
J. Am. Chem. Soc. 2005, 127, 11548–11549. (e) Rakow, N. A.; Suslick,
K. S. Nature 2000, 406, 710–713.
(3) Burg, S. P.; Burg, E. A. Plant Physiol. 1962, 37, 179–189.
(4) Lo´pez-Go´mez, R.; Campbell, A.; Dong, J. G.; Yang, S. F.; Go´mez-
Lim, M. A. Plant Sci. 1997, 123, 123–131.
(5) Klein, R. A.; Riley, M. R.; DeCianne, D. M.; Srinavakul, N. U.S.
Patent Appl. 127 543, 2006.
(2) McCoy, M.; Reisch, M. S.; Tullo, A. H.; Short, P. L.; Tremblay, J.-
F.; Storck, W. J. Chem. Eng. News 2007, 85 (27), 29–68.
(6) Kim, J.-H.; Shiratori, S. Jpn. J. Appl. Phys. 2006, 45, 4274–4278.
10.1021/ic800986t CCC: $40.75  2008 American Chemical Society
Inorganic Chemistry, Vol. 47, No. 19, 2008 9035
Published on Web 08/27/2008

Cabanillas-Galán et al.
1
received. H NMR spectra were recorded on a Bru¨ker Advance
DPX 300 or a Bru¨ker Advance DPX 500 at room temperature.
Chemical shifts (δ) are expressed in parts per million with a typical
dispersion of (0.01 ppm relative to tetramethylsilane for 1H NMR.
Coupling constants (J) are expressed in hertz with a typical
dispersion of (0.1 Hz. CHN elemental analyses and MS/FAB+
spectra were performed by the Analytical Services and Environ-
mental Projects at Queen′s University Belfast. IR spectra were
carried out on a Perkin-Elmer Sepctrum-RX or a Perkin-Elmer FTIR
Spectrum-One spectrometer, and samples were recorded as Nujol
dispersions held between polyethylene sheets (absorptions for Nujol:
2960-2840, 1455, and 1370 cm-1; absorptions for polyethylene:
728 and 718 cm-1). Solid state reactions were carried out in a Mixer
Mill type MM 200 from Retsh. Gas chromatography (GC) studies
were done using a 5890 Series II gas chromatograph with a porapak
column adapted using helium as the mobile phase. X-ray crystal
structures were obtained on a Bru¨ker AXS SMART 1000 at the
School of Chemistry and Chemical Engineering X-ray facility at
Queen′s University Belfast.
Figure 1. (a) Reversible coordination of ethylene by Fanizzi’s complexes,
[MX2(L1)],7 and (b) analogous compounds used in this work. Complexes
[PdCl2(L5)]11 and [PdCl2(L6)]10 have been reported previously.
Synthesis of Four-Coordinate Complexes. The ligands 2,2′-
dipyridyl, 1,10-phenanthroline, 2,9-dimethyl-1,10-phenanthroline
(L1), 2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline (bathocu-
proine, L5), and 6,6′-dimethyl-2,2′-dipyridyl (L6) were obtained
from Aldrich and used without further purification. The phenan-
throline ligands 2,9-di-n-butyl-1,10-phenanthroline (L2), 2,9-di-s-
butyl-1,10-phenanthroline (L3), and 2,9-diphenyl-1,10-phenanthro-
line (L4)9 as well as [PdCl2(L6)]10 and all complexes containing
L17 were prepared following reported procedures. Although the
structure of [PdCl2(L5)] has been reported,11 we have not found a
synthetic procedure in the literature, and its preparation is also
included here.
is specific to ethylene. Potassium permanganate is a strong
oxidizing agent which also reacts with alcohols and alde-
hydes. Pd(II)/molybdate mixtures are air-sensitive, and stable
sensor films are prepared through a complicated method
involving controlled deposition of the mixtures on layers of
silica nanoparticles and a polyelectrolyte.6 Thus, although a
number of compounds change color upon reaction with
ethylene, establishing a practical chemical basis for such
visual detecting still presents significant challenges in terms
of both sensitivity (e.g., maxima of 0.1-1 ppm are present
in fruit3) and selectivity (since oxygen and water as well as
other organic species may be present).
As a starting point toward meeting these challenges, we
used the work of Fanizzi et al.7,8 on the reversible binding
of olefins by air-stable platinum and palladium complexes
of 2,9-dimethylphenathroline (L1; Figure 1a). In particular,
they observed that compounds [MX2(L1)] (M ) Pd, Pt; X
) Cl, Br, I) coordinate ethylene reversibly in solution and
generally change color upon the formation of the trigonal-
bipyramidal ethylene complex (Figure 1a).7 The reaction
requires a planar bidentate ligand with bulky substituents
since steric hindrance around the metal center is then relieved
upon formation of the five-coordinate ethylene complex. In
order to prepare visual detectors that are practical and easy
to use, this solution chemistry had to be transferred to the
solid state, for example, by depositing the compounds on
solid supports. Here, the suitability of Fanizzi’s compounds
and of some new analogues for the visual detection of
ethylene is studied using a range of solid supports. The
chemical basis for a selective ethylene detector is described.
[PtCl2(L2)]. A solution of 2,9-di-n-butyl-1,10-phenanthroline
(L2; 1 g, 3.1 mmol) in CHCl3 (10.5 mL) was added to a solution
of K2[PtCl4] (1.289 g, 3.1 mmol) in water (10 mL). The mixture
was heated to reflux for 18 h. The organic fraction was separated
and the aqueous phase washed with CHCl3 (3 × 2 mL). The
combined organic fractions were dried by stirring with magnesium
sulfate, which was then removed by filtration. The filtrate was
concentrated to ca. 3 mL and added to cold diethyl ether (ca. 7
mL) while stirring vigorously. The product precipitated as an orange
solid and was collected by filtration. The filtrate was further
concentrated and treated with diethyl ether as described above. This
procedure was repeated three times to obtain the maximum yield.
All of the fractions of the solid thus obtained were combined,
washed with diethyl ether, and dried in the air. Yield: 1.219 g,
70%. 1H NMR (CDCl3): 0.93 [t, 6H, 3J(HH) ) 7.2 Hz, CH3], 1.38
[sextuplet, 4H, 3J(HH) ) 7.5 Hz, CH2], 1.85 [quintuplet, 4H, 3J(HH)
) 7.5 Hz, CH2], 3.72 [t, 4H, 3J(HH) ) 7.8 Hz, CH2], 7.55 [d, 2H,
3J(HH) ) 8.7 Hz, phenH(3, 8)], 7.78 [s, 2H, phenH(5, 6)], 8.39
3
[d, 2H, J(HH) ) 8.4 Hz, phenH(4, 7)] ppm. ν(Pt-Cl) ) 336
cm-1(br). Anal. calcd for C20H24N2Cl2Pt: C, 43.10; H, 4.31; N,
5.03%. Found: C, 43.00; H, 4.33; N, 5.01%. MS/FAB+: 541
(M+-CH3), 521 (M+-Cl), 499 (M+-C4H9), 485 (M+-2Cl), 293
(M+-2Cl, Pt).
Experimental Section
[PdCl2(L2)]. [PdCl2(NCPh)2] (0.223 g, 0.58 mmol) was dis-
solved in CHCl3 (130 mL). The solution was stirred and 2,9-di-n-
butyl-1,10-phenanthroline (L2; 0.268 g, 0.91 mmol) added. After
being stirred for 17 h, the solution was concentrated until becoming
General. Unless otherwise indicated, all starting materials and
volatile compounds were obtained from Aldrich and used as
(7) Fanizzi, F. P.; Intini, F. P.; Maresca, L.; Natile, G.; Lanfranchi, M.;
Tiripicchio, A. J. Chem. Soc., Dalton Trans. 1991, 1007–1015.
(8) (a) Fanizzi, F. P.; Maresca, L.; Natile, G.; Lanfranchi, M.; Tiripicchio,
A.; Pacchioni, G. J. Chem. Soc., Chem. Commun. 1992, 333–335. (b)
Fanizzi, F. P.; Natile, G.; Lanfranchi, M.; Tiripicchio, A.; Laschi, F.;
Zanello, P. Inorg. Chem. 1996, 35, 3173–3182.
(10) (a) Newkome, G. R.; Fronczek, F. R.; Gupta, V. K.; Puckett, W. E.;
Pantaleo, D. C.; Kiefer, G. E. J. Am. Chem. Soc. 1982, 104, 1782–
1783. (b) Newkome, G. R.; Pantaleo, D. C.; Puckett, W. E.; Ziefle,
P. L.; Deutsch, W. A. J. Inorg. Nucl. Chem. 1981, 43, 1529–1531.
(11) Rauterkus, M. J.; Krebs, B. Structure deposited in CCDC as private
communication (CCDC No. 193826), 2002.
(9) Dietrich-Buchecker, C. O.; Marnot, P. A.; Sauvage, J. P. Tetrahedron
Lett. 1982, 23, 5291–5294.
9036 Inorganic Chemistry, Vol. 47, No. 19, 2008

Detection of Ethylene Using Silica-Supported Pd Complexes
Figure 2. Crystal structures of (a) [PdCl2(L2)], (b) [PtCl2(L2)], (c) [PdCl2(L3)]·CHCl3, and (d) [PtCl2(C2H4)(L2)]. Hydrogen atoms and solvent molecules
have been omitted for clarity. Ortep representations (ellipsoids probability 30%) have been done using Ortep-3 for Windows.19
Table 1. Crystal Data
[PdCl2(L2)]
[PdCl2(L3)]·CHCl3
[PtCl2(L2)]
[PtCl2(C2H4)(L2)]
formula
M
C20H24Cl2N2Pd
469.71
C21H25Cl5N2Pd
589.08
C20H24Cl2N2Pt
558.40
C22H28Cl2N2Pt
586.45
crystal system
space group
a/Å
monoclinic
P21/n
orthorhombic
P212121
11.29(13)
12.04(11)
17.80(14)
2420(40)
298(2)
monoclinic
P21/n
triclinic
P1
j
13.7330(18)
10.2556(13)
14.5380(19)
1916.4(4)
153(2)
13.9102(18)
10.4746(14)
14.6778(19)
2010.8(5)
153(2)
9.0953(14)
10.9734(17)
12.0763(18)
1103.3(3)
153(2)
b/Å
c/Å
U/Å3
T/K
Z
4
4
4
2
µ(Mo KR)/mm-1
reflns measured
unique reflns (Rint
R (I > 2σ(I))
wR(F2) (all data)
1.252
21008
4331 (0.0356)
0.0386
0.1074
1.330
10794
4206(0.1288)
0.1215
0.3423
7.249
14839
3229(0.0536)
0.0448
0.1441
6.610
12114
4843(0.0634)
0.0581
0.1483
)
3
cloudy and diethyl ether (ca. 10 mL) added to give an orange solid.
CH], 7.58 [d, 2H, J(HH) ) 8.7 Hz, phenH(3, 8)], 7.86 [s, 2H,
3
The solid was collected by filtration, washed with diethyl ether,
phenH(5, 6)], 8.39 [d, 2H, J(HH) ) 8.4 Hz, phenH(4, 7)] ppm;
1
and dried in the air. Yield: 0.256 g, 93%. H NMR (CDCl3): 0.97
isomer 2, 0.75 [t, 6H, 3J(HH) ) 7.4 Hz, CH3], 1.37 [d, 6H, 3J(HH)
) 6.9 Hz, CH3], 1.88-1.66 [m, 2H, CH2], 4.35 [sextuplet, 2H,
[t, 6H, 3J(HH) ) 7.4 Hz, CH3], 1.42 [sextuplet, 4H, 3J(HH) ) 7.4
3
3
Hz, CH2], 1.86 [quintuplet, 4H, J(HH) ) 7.7 Hz, CH2], 3.71 [t,
3J(HH) ) 8.0 Hz, CH], 7.58 [d, 2H, J(HH) ) 8.4 Hz, phenH(3,
4H, 3J(HH) ) 8.0 Hz, CH2], 7.55 [d, 2H, 3J(HH) ) 8.4 Hz,
8)], 7.85 [s, 2H, phenH(5, 6)], 8.38 [d, 2H, J(HH) ) 8.4 Hz,
3
3
phenH(3, 8)], 7.85 [s, 2H, phenH(5, 6)], 8.35 [d, 2H, J(HH) )
phenH(4, 7)] ppm. ν (Pd-Cl) ) 326 cm-1 (br). Anal. calcd for
C20H24N2Cl2Pd: C, 51.14; H, 5.15; N, 5.96%. Found: C, 51.07; H,
5.27; N, 5.94%. MS/FAB+: 437 (M+-Cl), 415 (M+-C4H9), 397
(M+-2Cl), 293 (M+-2Cl, Pd).
8.4 Hz, phenH(4, 7)] ppm. ν(Pd-Cl) ) 336 cm-1 (br). Anal. calcd
for C20H24N2Cl2Pd·1/5CHCl3: C, 49.26; H, 4.91; N, 5.69%. Found:
C, 49.34; H, 4.68; N, 5.81%. MS/FAB+: 453 (M+-CH3), 433
(M+-Cl), 411 (M+-C4H9), 397 (M+-2Cl), 293 (M+-2Cl, Pd).
[PdCl2(L3)]. To a solution of [PdCl2(NCPh)2] (0.05 g, 0.13
mmol) in CHCl3 (5 mL) was added 2,9-di-s-butyl-1,10-phenan-
throline (L3; ca. 0.02 g, ca. 0.07 mmol). The solution was stirred
for 19 h and filtered through Celite. The filtrate was concentrated
to ca. 1.5 mL and added dropwise to cold diethyl ether (ca. 3 mL)
with vigorous stirring. A brown solid precipitated, which was
collected by filtration, washed with diethyl ether, and dried in the
air. Yield: 0.011 g, ca. 32%. 1H NMR (CDCl3): isomer 1, 0.97 [t,
6H, 3J(HH) ) 7.4 Hz, CH3], 1.53 [d, 6H, 3J(HH) ) 6.9 Hz, CH3],
[PtCl2(L4)]. To a stirred suspension of [PtCl2(dmso)2] (0.133
g, 0.32 mmol) in CHCl3 (35 mL) was added 2,9-diphenyl-1,10-
phenanthroline (L4; 0.110 g, 0.33 mmol). The suspension was
heated to reflux for 24 h. The solid dissolved after 40 min, and the
resulting solution became orange after a further 20 min. A
suspension formed after 90 min at reflux. The solid was collected
by filtration, washed with diethyl ether, and dried in the air. Yield:
0.155 g, 81%. 1H NMR (dmso-d6): 7.20-7.09 [m, 2H, phen],
7.70-7.45 [m, 5H, phen], 8.20-7.72 [m, 3H, phen+Ph], 8.27 [d,
1H, 3J(HH) ) 8.7 Hz, phen], 8.43 [d, 1H, 3J(HH) ) 8.4 Hz, phen],
8.51 [d, 1H, 3J(HH) ) 7.2 Hz, Ph], 8.62 [d, 1H, 3J(HH) ) 8.4 Hz,
3
1.88-1.66 [m, 2H, CH2], 4.36 [sextuplet, 2H, J(HH) ) 8.0 Hz,
Inorganic Chemistry, Vol. 47, No. 19, 2008 9037

Cabanillas-Galán et al.
Table 2. Selected Bond Distances (Å) and Angles (deg)
[PdCl2(L3)]·
containing ethylene. The solution was stirred for 22 h. Diethyl ether
was added to the solution to form a yellow solid, which was
collected by filtration, washed with diethyl ether, and dried in the
[PtCl2(C2H4)
[PdCl2(L2)]
CHCl3
[PtCl2(L2)]
(L2)]
1
3
air. Yield: 0.008 g, 8%. H NMR (CDCl3): 1.07 [t, 6H, J(HH) )
M-Cl1
2.2779(8)
2.2944(8)
2.074(3)
2.051(2)
2.258(17)
2.259(17)
1.97(2)
2.340(3)
2.330(3)
2.080(9)
2.082(9)
2.328(2)
2.318(2)
2.266(8)
2.242(7)
2.073(8)
2.086(10)
1.464(14)
88.38(17)
90.49(18)
178.85(8)
74.6(3)
92.54(17)
89.10(18)
41.2(4)
90.0(3)
89.8(3)
3
7.5 Hz, CH3], 1.63 [sextuplet, 4H, J(HH) ) 7.8 Hz, CH2], 2.07
M-Cl2
3
3
M-N1
[quintuplet, 4H, J(HH) ) 7.8 Hz, CH2], 3.91 [t, 4H, J(HH) )
M-N12
Pt-C23
Pt-C24
C23-C24
N1sM-Cl1
1.96(3)
8.1 Hz, CH2], 3.66 [m, 4H, CH2dCH2, J(195Pt-H) ) 69 Hz], 7.82
3
[d, 2H, J(HH) ) 8.4 Hz, phenH(3,8)], 7.86 [s, 2H, phenH(5,6)],
8.36 [d, 2H, 3J(HH) ) 8.1 Hz, phenH(4,7)] ppm. ν (Pt-Cl) ) 340
cm-1. Anal. calcd for C22H28N2Cl2Pt: C, 45.06; H, 4.81; N, 4.78%.
Found: C, 44.54; H, 4.67; N, 4.57%. MS/FAB+: 587 (M+), 551
(M+-Cl), 521 (M+-C2H4Cl), 485 (M+-C2H4, 2Cl).
170.41(7)
97.5(9)
92.6(9)
87.1(8)
82.1(10)
162.0(6)
177.7(7)
98.3(2)
94.5(2)
85.98(9)
80.3(3)
172.1(3)
171.1(2)
N12sM-Cl2 169.45(7)
Cl1-M-Cl2
N1-M-N12
N1-M-Cl2
N12-M-Cl1
C23-Pt-C24
C23-Pt-Cl2
C24-Pt-Cl2
86.76(3)
80.62(9)
97.08(6)
94.00(7)
Crystal Structures. Single crystals of [MCl2(L2)] (M ) Pt, Pd),
[PdCl2(L3)]·CHCl3, and [PtCl2(C2H4)(L2)] suitable for X-ray
studies were obtained by the slow evaporation of CHCl3 or CH2Cl2
solutions of the compounds. Each crystal was mounted onto the
diffractometer at low temperatures under nitrogen at ca. 120 K (253
K for [PdCl2(L3)] ·CHCl3). The structures were solved using direct
methods and refined with SHELXTL.12 The non-hydrogen atoms
were refined with anisotropic thermal parameters. Hydrogen-atom
positions were added at idealized positions with a riding model
and fixed thermal parameters (Uij ) 1.2Ueq for the atom to which
they are bonded). The function minimized was ∑[w(|Fo|2 - |Fc|2)]
with reflection weights w-1 ) [σ2 |Fo|2 + (g1P)2 + (g2P)] where P
) [max|Fo|2 + 2|Fc|2]/3. Additional material available from the
Cambridge Crystallographic Data Centre comprises relevant tables
of atomic coordinates, bond lengths and angles, and thermal
parameters.
Experiments on Solid Supports. Preparation of Detector
Plates. Six types of supports were used for the experiments: silica
gel 60F254 on aluminum obtained from Merck; reverse-phase silica
plates, KC2F ethyl reverse-phase, and K2F cellulose obtained from
Whatman; cellulose on polyester obtained from Aldrich; and
Polygram Alox N/UV254 obtained from Macherey-Nagel. To ensure
consistency, reference color cards were used to monitor any changes
of color. All of the detectors were prepared following the same
procedure: Solutions of each metal complex were prepared in two
concentrations, 2.5 and 5 mM, in CHCl3. In the case of [PtCl2(L4)],
a few drops of dimethylsulfoxide were added to dissolve the solid
completely. Seven drops of each solution (ca. >0.24 mL) were
applied to the support to create a spot. The plate was placed in a
desiccator jar in an oil bath and dried under a vacuum at 50 °C
over 1 h. The dry plates were stored under a vacuum until used.
At this point, the colors of the spots were similar to those of the
corresponding complexes and ranged from yellow to light-brown
or gray/green.
3
3
phen], 8.77 [d, 1H, J(HH) ) 8.4 Hz, phen], 8.86 [d, 1H, J(HH)
) 8.7 Hz, phen] ppm. ν (Pt-Cl) ) 304 and 341 cm-1. Anal. calcd
for C24H16N2Cl2Pt: C, 48.17; H, 2.69; N, 4.68%. Found: C, 47.85;
H, 3.25; N, 4.46%. MS/FAB+: 562 (M+-Cl), 526 (M+-2Cl), 333
(M+-2Cl, Pt).
[PdCl2(L5)]. Bathocuproine (L5; 0.08 g, 0.2 mmol) was added
to a solution of [PdCl2(NCPh)2] (0.09 g, 0.2 mmol) in CHCl3 (25
mL) while stirring. After being stirred for 20 h, the solution was
filtered through Celite. The filtrate was concentrated to ca. 5 mL,
and diethyl ether was added to precipitate an orange solid, which
was collected by filtration, washed with diethyl ether, and dried in
the air. Yield: 0.102 g, 86%. 1H NMR (CDCl3): 3.30 [s, 6H, CH3],
7.48-7.82 [m, 14H, phen+Ph] ppm. ν (Pd-Cl) ) 304, 342 cm-1
.
Reactions with Ethylene in Solution. Each complex (ca. 0.02
mmol) was dissolved in ca. 3.5 mL of CDCl3 or dmso-d6
([PtCl2(L4)]) and placed in a flask. Ethylene gas was bubbled
directly into the solution through a needle while it was stirred for
90 min. Significant changes of color were observed only in the
case of L1 derivatives, in agreement with a study previously
reported.7 The H NMR spectra of the resulting solutions were
1
taken. Complete conversion to the corresponding five-coordinate
complex was observed for L1 complexes7 and [PtCl2(L2)]. The
ethylene derivative of the latter was isolated and fully characterized
(see below). Partial coordination occurred for [PdCl2(L2)] and
[PdCl2(L3)], and no reaction was observed for any of the complexes
containing L4-L6. The 1H NMR data for the ethylene derivatives
of [PdCl2(L2)] and [PdCl2(L3)], as shown by the NMR spectra of
the mixtures, are as follows.
1
3
[PdCl2(C2H4)(L2)]. H NMR (CDCl3): 1.09 [t, 6H, J(HH) )
Control Experiments. To monitor any changes of color of the
detector plates toward water, each plate was placed on a frit above
water (ca. 20 mL) in a closed desiccator jar for 14 days. To monitor
any changes of color due to the air, the plates were left open to the
air for 14 days. Changes of color were monitored every 24 h.
Exposure to Ethylene or Propene. Each plate was placed in a
Schlenk tube. The air was removed, and an ethylene- or propene-
filled balloon was connected. The gas atmosphere was maintained
for one week, and the color changes were monitored approximately
every hour during the first 12 h, and then every 24 h. Most plates
significantly changed color within the first 24 h, with no further
changes observed afterward.
3
7.5 Hz, CH3], 1.68 [sextuplet, 4H, J(HH) ) 7.5 Hz, CH2], 2.14
3
3
[quintuplet, 4H, J(HH) ) 7.8 Hz, CH2], 4.00 [t, 4H, J(HH) )
8.2 Hz, CH2], 4.94 [s, 4H, CH2dCH2], 7.77 [d, 2H, 3J(HH) ) 8.5
Hz, phenH(3,8)], 7.82 [s, 2H, phenH(5,6)], 8.32 [d, 2H, 3J(HH) )
1
8.5 Hz, phenH(4,7)] ppm. [PdCl2(C2H4)(L3)]: H NMR (CDCl3):
isomer 1, 0.95 [t, 6H, 3J(HH) ) 7.4 Hz, CH3], 1.52 [d, 6H, 3J(HH)
) 6.6 Hz, CH3], 1.95-1.58 [m, 2H, CH2], 4.32 [sextuplet, 2H,
CH], 4.92 [s, 4H, CH2dCH2], 7.57 [d, 2H, 3J(HH) ) 8.7 Hz,
phenH(3,8)], 7.86 [s, 2H, phenH(5,6)], 8.41 [d, 2H, 3J(HH) ) 8.7
3
Hz, phenH(4,7)] ppm; isomer 2, 0.73 [t, 6H, J(HH) ) 7.2 Hz,
3
CH3], 1.35 [d, 6H, J(HH) ) 6.9 Hz, CH3], 1.95-1.58 [m, 2H,
CH2], 4.32 [sextuplet, 2H, CH], 4.92 [s, 4H, CH2dCH2], 7.56 [d,
Exposure to Volatile Liquids. The plates were placed in a
Schlenk tube (open to the atmosphere). In an adjacent Schlenk tube,
equipped with a rubber septum, the corresponding volatile was
2H, 3J(HH) ) 8.7 Hz, phenH(3,8)], 7.85 [s, 2H, phenH(5,6)], 8.40
3
[d, 2H, J(HH) ) 8.7 Hz, phenH(4,7)] ppm.
[PtCl2(C2H4)(L2)]. A solution of [PtCl2(L2)] (0.098 g, 0.18
mmol) in CHCl3 (4 mL) was prepared in a Schlenk tube and stirred.
The air was removed and the tube connected to a rubber balloon
(12) Sheldrick, G. SHELXTL, version 6.10; Bru¨ker AXS: Madison, WI,
2000.
9038 Inorganic Chemistry, Vol. 47, No. 19, 2008

Detection of Ethylene Using Silica-Supported Pd Complexes
placed, and the Schlenk tubess were connected Via a rubber tube.
A dinitrogen flow was bubbled into the volatile liquid using a needle
through the septum and passed through the Schlenk tube containing
the plate. The flow was maintained for 1 h, during which color
changes were monitored.
Experiments with Fruit. Silica plates spotted with a 5 mM
solution of [PdCl2(L2)] in chloroform, as described above, were
placed in desiccator jars containing bananas. In the first experiment,
green bananas in a very early ripe stage were used. The detector
plates exposed to these changed color from orange to green/gray
after 4 days. In a second experiment, already ripened yellow bananas
were used. In this case, the bananas were kept for 12 days, until
rotten. After this time, only a very slight darkening of the detector
plates was observed.
Monitoring of Ethylene Production Using Gas Ghromatog-
raphy. Ethylene at a concentration of 10 ppm was used as a
reference. Volumes of 0.1-0.5 mL were taken directly from an
ethylene cylinder at two different pressures (0.5 and 1 bar) and
injected into a GC Via syringe. The retention times for the reference
ethylene samples were in the range 0.630-0.646 min. Bananas and
a silica-[PdCl2(L2)] detector plate were introduced in a headspace
with a septum adapter in the upper part. The container was closed
and left to equilibrate for 2 h. Samples (0.5 mL) were taken daily
through the septum and injected into the GC. In a first experiment,
three green, unripe bananas were used. The bananas remained green
for 7 days. During this time, no ethylene was detected by GC, and
the detector plate did not change color significantly. In a second
experiment, one yellow and two unripe bananas were left in the
headspace for 4 days. Ethylene was detected by GC every day from
the first day, and the detector plate clearly changed color after 2
days. A second plate was introduced on the third day that also
changed color after 24 h.
Figure 3. Crystal packing for [PtCl2(L2)] showing its layered structure
with molecules connected through π-π (3.22, 3.35 Å) and Cl · · ·H-C
interactions (Cl1 · · ·H6A, 2.837 Å; Cl1-C6, 3.734 Å; Cl1 · · ·H6A-C6,
158°; Cl2· · ·H3A, 2.815 Å; Cl2-C3, 3.688 Å; Cl2 · · ·H3A-C3, 152°).
the expected distortions arising from the bulky chelating
phenanthroline ligand, in particular, the bending of the
phenanthroline backbone and the displacement of the Cl
atoms from the N-metal-N plane.7,8 Complexes [MCl2(L2)]
(M ) Pt, Pd) form layered structures in which layers of
interdigitated n-Bu chains alternate with layers of π-π
stacked (3.2-3.4 Å) phenanthroline groups. A number of
inter- and intralayer CH · · ·Cl hydrogen-bonding interactions
are also present (see Figure 3 and the Supporting Informa-
tion). In contrast, complex [PdCl2(L3)]·CHCl3 does not form
layers but exhibits an extended H-bonded structure involving
the solvent molecules (see the Supporting Information). The
1H NMR spectrum of [PdCl2(L3)] shows a ca. 1:1 mixture
of the RR/SS and RS/SR diastereomers, although the crystal
used for X-ray studies shows only the meso (RS) isomer.
Reactions with Ethylene in Solution. The reactions of
the complexes with excess ethylene in solution were
Results and Discussion
Synthesis and Characterization of Four-Coordinate
Complexes. In addition to Fanizzi’s complexes (Figure 1a),7
a series of new four-coordinate platinum and palladium
compounds containing 2,9-disubstituted phenanthroline ligands
(L2-L5, Figure 1b) were prepared. For comparison, the Pd
derivative of the more flexible 2,2′-dimethyl-6,6′-dipyridyl,
[PdCl2(L6)],10 was also included in the study (Figure 1b).
The only unusual features with regard to the spectroscopic
1
data of these compounds are found in [PtCl2(L4)]. Its H
NMR spectrum in dmso-d6 shows asymmetry of the phenan-
throline ligand, which may be due to restricted rotation in
solution of the bulky Ph substituents, and potentially the
formation of a Cl · · · H-C(Ph) interaction. All complexes
show in their IR spectra either one broad band or two bands
in the region 325-345 cm-1, compatible with a cis-MCl2
(M ) Pt, Pd) geometry13 (see the Experimental Section and
Supporting Information). However, [PtCl2(L4)] exhibits an
additional low-frequency band at 304 cm-1, which supports
the involvement of one of the Cl atoms in H bonding.14
The crystal structures of three new phenathroline com-
plexes, [MCl2(L2)] (M ) Pt, Pd) and [PdCl2(L3)]· CHCl3,
were obtained (Figure 2a-c, Tables 1 and 2). They all show
1
monitored by H NMR spectroscopy. Complete conversion
into the five-coordinate species was observed for Fanizzi’s
L1 complexes, as previously reported,7 and for [PtCl2(L2)],
whereas only partial coordination of ethylene was observed
in the cases of [PdCl2(L2)] and [PdCl2(L3)]. However,
significant changes of color were observed only for CDCl3
solutions of [PdCl2(L1)] (orange to yellow), [PtBr2(L1)] (red
to orange), and [PdI2(L1)] (gray to brown).
The 1H NMR resonances corresponding to the five-
coordinate derivatives [MCl2(C2H4)(L2)] (M ) Pt, Pd) and
[PdCl2(C2H4)(L3)] were similar to those of the analogous
L1 complexes,7 with coordinated ethylene appearing at ca.
4.9 ppm for [PdCl2(C2H4)(L2)] or [PdCl2(C2H4)(L3)], and
at 3.66 ppm [J(195Pt-H) ) 69 Hz] for [PtCl2(C2H4)(L2)].
Incomplete conversion into the ethylene complex in the
cases of [PdCl2(L2)] and [PdCl2(L3)] is not unexpected, since
(13) Sakai, K.; Tomita, Y.; Ue, T.; Goshima, K.; Ohminato, M.; Tsubomura,
T.; Matsumoto, K.; Ohmura, K.; Kawakami, K. Inorg. Chim. Acta
2000, 297, 64.
(14) Spaniel, T.; Gorls, H.; Scholz, J. Angew. Chem., Int. Ed. 1998, 37,
1862–1865.
Inorganic Chemistry, Vol. 47, No. 19, 2008 9039

Cabanillas-Galán et al.
Figure 5. Silica plate impregnated with (a) [PdCl2(L2)] or (b) [PtCl2(L2)]
before and after exposure to ethylene gas. (c) Proposed mechanism for the
formation of Pd black and the darkening of the sensor in the presence of
ethylene.
Figure 4. Crystal packing for [PtCl2(C2H4)(L2)] showing π-π stacking
(i.e., shortest π-π distance is ca. 3.45 Å).
color for longer periods when exposed to the air or water
vapor than those prepared using 5 × 10-3 M solutions.
However, the lower loadings also tended to show less
significant color changes upon exposure to ethylene. In
general, the platinum derivatives were more affected by water
and/or air than the palladium analogues. Although further
research is needed to fully understand the role of the support,
it is interesting to note that the response to ethylene was
specific to the type of support used, with selective responses
to ethylene observed mainly on silica or reverse-phase silica
preparations (see the Supporting Information for a summary
of the color changes observed).
From the above studies, various complex-support com-
binations were identified as promising detectors; that is, the
detector’s color was visually stable when exposed to the air
and water but gave an easily detectable change of color upon
exposure of ethylene. From these, [PdCl2(L2)] on silica was
chosen as the optimum system for further studies because it
gave one of the clearest color changes (from very light orange
to green, Figure 5a) at a low loading within 1 h, and because
of the high solubility and ease of preparation of the metal
complex itself.
It is likely that the mechanism of the reduction involves a
Wacker-type oxidation of ethylene in the presence of traces
of water (Figure 5c). However, due to the small quantities
of product released against the large ethylene background
concentration, and the extended period over which the
reaction takes place, it has not been possible to confirm this
by identifying the byproduct of the reaction.
The first step of a Wacker process is the formation of an
ethylene-Pd(II) complex, which in the present case may
involve a species analogous to the five-coordinate species
formed in solution. In contrast to that observed in the
solid-gas reaction, decomposition to Pd black has not been
observed for the five-coordinate ethylene-Pd(II) complexes
in solution, but this may be due to the lability of the olefin
ligand under these conditions (see above). Further work is
planned in order to identify the intermediates and fully
understand the mechanism of the reaction between the
supported complexes and ethylene.
olefin coordination has been shown to be less favored in
palladium than in platinum complexes.7 Even in the presence
1
of a large excess of ethylene, the H NMR spectra of these
mixtures showed four- and five-coordinate species together
with free ethylene. The latter appeared as an intense singlet
at 5.4 ppm. In the case of [PdCl2(C2H4)(L3)], signals
corresponding to the RR/SS and RS/SR diastereomers were
also identified.
Derivatives with L4 or L5 did not react with ethylene,
presumably due to the electron-withdrawing phenyl substit-
uents. In addition, no reaction with ethylene was observed
for the more flexible dipyridyl derivative [PdCl2(L6)].
The crystal structure of [PtCl2(C2H4)(L2)] was obtained
(Figure 2d, Tables 1 and 2) and shows the expected trigonal-
bipyramidal geometry at the metal center, with the equatorial
positions occupied by the nearly planar phenanthroline ligand
and the olefin. Extended π-π stacking involving the
phenanthroline moieties is also observed (Figure 4).
Reactions of Supported Complexes with Ethylene. To
investigate ethylene coordination in the solid state, samples
of each compound were deposited on a variety of solid
supports, including silica, reverse-phase silica, cellulose, and
alumina. The potential detectors were prepared by depositing
ca. 0.24 cm3 of a chloroform solution (2.5 × 10-3 or 5 ×
10-3 M) of each compound over each of the solid supports.
The plates were then dried and exposed to ethylene gas. Most
complexes underwent significant changes of color within a
few hours of exposure to ethylene, and on more than one
type of support. Interestingly, there was no direct correlation
with their reactivities in solution. The changes were irrevers-
ible and generally consisted of darkening to brown or dark
gray, suggesting partial decomposition to palladium or
platinum black (Figure 5a,b). Control experiments to assess
the stability of the samples toward air or water vapor were
also performed. During these control tests, color changes
were observed in many cases, but the final colors were often
different from those obtained upon ethylene exposure. For
example, the colors of the plates generally became lighter
when left in the air. The plates prepared using 2.5 × 10-3
M solutions of the complexes generally kept their original
To assess the influence of the bidentate ligand on the
response to ethylene, the derivatives containing nonsubsti-
9040 Inorganic Chemistry, Vol. 47, No. 19, 2008

Detection of Ethylene Using Silica-Supported Pd Complexes
lene, and p-tolylacetylene (Figure 6). However, these other
analytes are not relevant in the context of fruit ripening or
ethylene pipe-leak detection because they are not expected
to be present at significant amounts. Moreover, other fruit
volatiles, such as hexylalcohol, 3-methyl-1-butanol, and
2-pentanone did not produce a color change. This is
important because alcohols and carbonyl compounds are also
emitted by climacteric fruit at various stages but may not be
reliable indicators of ripening.17
Figure 6. Color changes observed for [PdCl2(L2)]-silica detectors upon
exposure to p-tolylacetylene and phenylacetylene.
tuted ligands [PdCl2(1,10-phenanthroline)] and [PdCl2(2,2′-
dipyridyl)]15 were also tested after deposition on silica,
together with [PdCl2(NCPh)2]. Upon exposure to ethylene,
air, or water, neither of the former two complexes underwent
significant color changes, whereas [PdCl2(NCPh)2] darkened
significantly in all cases. From this, it appears that the
complexes with nonsubstituted bidentate ligands are too
stable and are unresponsive, while in the absence of a
stabilizing chelating ligand, [PdCl2(PhCN)2] very quickly
decomposes to Pd black. With 2,9-disubstituted bidentate
ligands, the complexes are sufficiently stable toward water
and air, but still reactive toward ethylene.
We extended the investigation to see if detectors of this type
would respond differently to fruit at different stages of ripeness.
Experiments with fruit were performed with [PdCl2(L2)]-silica.
When placed in a sealed vessel with green, preclimacteric
bananas for 24 h, a similar color change to that seen upon
exposure to ethylene occurred (light-orange to green/gray).
Significantly, when exposed to already-ripe bananas under the
same conditions, no response was observed. These results are
consistent with the known emission profile for ethylene from
climacteric fruit, in which an ethylene “spike” is observed during
the climacteric itself, from which point ripening accelerates.3,16
In complementary experiments, ethylene production from the
fruit was monitored by GC. No ethylene was detected from
the green bananas, and detector plates placed next to the fruit
did not change color. However, when a yellow banana was
added to the experiment to induce ripening, ethylene was
detected by GC at 24 and 48 h after addition of the yellow
banana, and a clear color change of the detector was observed
after 2 days.
Reactions of Supported Complexes with Other Vola-
tiles. In order to assess the selectivity of the detectors, the
responses of silica-supported [PdCl2(L2)] or [PtCl2(L2)]
toward other organic volatiles were studied. [PtCl2(L2)]
underwent significant darkening (to brown) in most cases,
whereas the response of the Pd detector was more selective.
Thus, [PdCl2(L2)]-silica detectors did not change color upon
exposure to alkenes such as 1-decene, 1-dodecene, styrene,
cyclopentene, cyclooctene, 1,3-butadiene, or acrylonitrile.
Slight color changes were observed upon exposure to
aromatic compounds (toluene, pyridine, and benzonitrile) and
to some terminal alkynes, such as 2-butyne and (4-pen-
tylphenyl)acetylene, whereas significant darkening to brown
or red was observed upon exposure to propene, phenylacety-
Conclusions
In summary, we report a simple system consisting of a
palladium complex on a silica support which shows a clear
visible response to ethylene and, therefore, represents a
chemical basis for visual ethylene detection that can be
applied in pipe-leak detection and the monitoring of fruit
ripening. Although the detectors have also been found to
respond to propene and terminal aromatic acetylenes, these
volatiles are not significant in the context of the mentioned
applications. Encouragingly, discrimination between fruit at
different stages of ripening has also been observed. Overall,
[PdCl2(L2)]-silica detectors are easy to prepare and use,
and they show improved specificity and stability compared
with previous visual sensors (i.e., those based on perman-
ganate or Pd(II)/molybdate). It is interesting to note that the
color change (to red) upon exposure to p-tolylacetylene is
specific for this volatile.
With regard to cost, it should be noted that an effective
detector requires only 0.5-1 mg of [PdCl2(L2)]; that is,
starting from 1 g of PdCl2, it is possible to prepare over 2000
detectors, and therefore cost is unlikely to be prohibitive.
The system can clearly also be varied with regard to the types
of coligands present, so that the visible response may be
tuned. In addition, color-changing detectors of this type can
be incorporated into low-cost optical devices that allow
remote wireless detection,18 with important potential ap-
plications including the detection of ethylene leaks from
pipelines. Work is currently underway in this area.
Acknowledgment. We are grateful to the European Social
Fund and the Questor Centre, Belfast, for funding.
Supporting Information Available: Summary of color changes
upon exposure to ethylene and control experiments, IR data,
additional figures of the crystal packing of complexes [PdCl2(L2)]
and [PdCl2(L3)]·CHCl3. This material is available free of charge
via the Internet at http://pubs.acs.org.
IC800986T
(17) (a) Shiota, H. J. Agric. Food Chem. 1993, 41, 2056–2062. (b) Aubert,
C.; Gunata, Z.; Ambid, C.; Baumes, R. J. Agric. Food Chem. 2003,
51, 3083–3091. (c) Bicalho, B.; Pereira, A. S.; Neto, F. R. A.; Pinto,
A. C.; Rezende, C. M. J. Agric. Food Chem. 2000, 48, 1167–1174.
(d) Visai, C.; Vanoli, M. Sci. Hortic. 1997, 70, 15–24. (e) Chapman,
G. W.; Horvat, R. J.; Forbus, W. R. J. Agric. Food Chem. 1991, 39,
867–870.
(18) (a) Shepherd, R. L.; Yerazunis, W. S.; Lau, K. T.; Diamond, D. IEEE
Sens. J. 2006, 6, 861–866. (b) Lau, K. T.; Yerazunis, W. S.; Shepherd,
R. L.; Diamond, D. Sens. Actuators, B 2006, 114, 819–825.
(19) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(15) Pazderski, L. Magn. Reson. Chem. 2006, 44, 163–170.
(16) Liu, X. J.; Shiomi, S.; Nakatsuka, A.; Kubo, Y.; Nakamura, R.; Inaba,
A. Plant Physiol. 1999, 121, 1257–1265.
Inorganic Chemistry, Vol. 47, No. 19, 2008 9041