In this project we investigated the use of Surface Enhanced Raman Spectroscopy (SERs) to detect and quantify Voltile Organic Compounds and Organophosphates in vapour and liquid phases, using a combination of temperature controlled flow cell and highly integrated and custom spectrometer using a Digital Mirror Device. The goal was the development of a low cost raman spectroscopy system that could be rapidly and widely deployed to detect volitile organic compounds and organophosphates. We used tricresyl phosphates (TCP) as an example VOC which is of interest to the aircraft industry and the chemical warfare simulatants dimethyl methyl phosphonate (DMMP) and di-isopropyl methyl phosphonate (DIMP) which have interests relating to counter terrorism applications. The project was to work towards producing a sensing system that could be deployable and automatically collect and analyse vapour and liquid samples.

Key Points

  • Detecting & Quantifying VOC's & Organophosphates in vapour at human health relevant concentrations.

  • Analytes considered: BPE (for benchmarking), tricresyl phosphates (TCP), dimethyl methyl phosphonate (DMMP), and di-isopropyl methyl phosphonate (DIMP).

  • Investigating & Comparing Surface Enhanced Substrates from Nanoparticles to electrochemical rough gold.

  • Using electrochemical methods to clean and replenish the substartes for continuous service.

  • Design & Development of a temperature controlled flow cell for condensating the vapour.

  • Working with binary and masking mixtures.

  • Design & Development of highly integrated Digital Mirror Device (DMD) spectrometer.

  • Benchmarking off-the-shelf Raman Spectrometer with custom DMD spectrometer.

  • Python based software for system automation and spectrograph analysis.

Experimental System


spectroscopy electrochemistry python microfluidics


External Links & Publications

Vapour Phase Compounds

From the literature it was deemed that dimethyl methyl phosphonate (DMMP), di-isopropyl methyl phosphonate (DIMP) and should be measured with a limit of detection of 10 μg mL−1 and 2 μg mL−1, respectively. Tricresyl phosphates (TCP) has a span of reported concentrations, the mean stated being 0.07 μg m-3. We created the apparatus to vapourise these compounds and flow them at known concentrations over our flow cell, which were then condensed onto the measurement substrate.

Substrates, Preparation & Electrochemical Cleaning

We analysed 4 different types of substrates. We investigated gold nanoparticles drop cast, a paper based nanoparticle SERS substrate, a nanotechnology gold substrate from Q-SERS and sreen printed Gold electrode. The idea behind all were that they could be integrated into our flow cell (with the expection of the drop cast) and all could act as a working electrode. The SER substrate provided the Raman interaction required, however, we were interested in how to use the same electrode for subsequent assays and for this we investigated electrohemical methods of cleaning to remove the bound compounds and replenish the SERS, this is an interesting technique so please contact Platform Kinetics for more information.

Experimental Setup

The experiments were carried out at 785nm wavelengths. The laser was a LuxxBox which was controlled via our Python software. This was fibre coupled to a OceanOptics RPB785 probe and with Z height adjustable mechanics to control the focus onto the substrate. The spectrometer used was an OceanOptics QE-Pro-Raman with shift range 0-2800 cm-1 and a mean resolution of 8 cm-1.
In real world situations sample matrices do not contain only a single sample. Multiple samples can mask one another and therefore are hard to differentiate. In this application we investigate the difference in compounds boiling points to seperate the compounds on the substrate by reversing the condensation mechanism on the flow cell, this is patent pending technique, please contact Platform Kinetics for more information.

Custom Spectrometer & Electronics

The design of a low-cost, custom Raman Spectrometer was investigated using a Digital Mirror Device from Texas Instruments. The idea behind this approach was to increase the sensitivity and reduce the optical part count. By using a DMD the linear array could be replaced with a highly sensitive InGaAs Detector, please contact Platform Kinetics for more information.


The software was based on Python which was used to control the automation and sequencing of the apparatus using a USB connection. The resulting raman spectra was collected and analysed. Initially the collected data was filtered, before being processed. Several technqiues were employed such as peak detection and wavelet transforms. The main requirement was to a) identify the compound(s) and b) quantify the concentration of the compound. The methods of identification came from training an machine learning algorithm based on a method called sembalance. The result was a closed loop system that could operate, collect and condense samples, measure, clean, identify and repeat the process
Software like this can be applied to your experimental set-ups, please contact Platform Kinetics for more information.