Abstract
Modern mining industry undergoes extensive drilling campaigns during exploration, mining, processing and
recycling of raw materials. There is a growing demand for more precise and automated digital processes to reduce
the needed time, cost and labour of ore exploration and handling. A precise and analytical drill core logging
process utilising multi-sensor approach could provide fast and timely method for mineralogy tracking. Most of the
conventional spectroscopic methods in drill core analysis reveal their elemental information. This information can
be used to create models and predict the mineralogy, but it does suffer from the inability to detect the molecular
bonds and autonomously build precise predictions. Elemental information requires background information on the
sample mineralogy to be able to build the models and it can be disturbed by similar mineral consistencies. Raman
spectroscopy is a good support for data fusion with elemental spectroscopy such as XRF and LIBS, because it can
indicate the molecular bonds instead of elements [1]. However, the conventional Raman spectrometers have limited
amount of minerals they can distinguish, because the fluorescence of the sample can lower the detection sensitivity
or sometimes overflow the signal entirely. This problem is common especially in minerals. A solution for reducing the
elevated signal background comes in the form of time-gated Raman spectroscopy [2]. This technology uses short laser
pulses to excite the sample and measures the scattered radiation in short time-window before the slower timescale
phenomena, fluorescence, has time to manifest. In many cases, this can eliminate the fluorescence background
and bring on to view spectral features that are commonly hidden or weak using traditional Raman techniques. In
this study, timegated PicoRaman M2 spectrometer, utilizing 532 nm pulsed excitation laser, is integrated to a multi-
sensor Ancorelog drill core logger solution as a mineral identification tool. The measurement utilises an auto-focus
probe developed at VTT with ability to set the system in focus autonomously during measurements. In addition, a
mineral identification software is developed for automated identification of minerals using Raman spectroscopy. The
solution utilizes reference spectra from the RRUFF online database of minerals [3]. A classification model correlates
the measured spectra with reference spectra, which are selected using expert knowledge of the sample mineralogy.
The correlation utilizes a dot product between the spectra as a metric.
recycling of raw materials. There is a growing demand for more precise and automated digital processes to reduce
the needed time, cost and labour of ore exploration and handling. A precise and analytical drill core logging
process utilising multi-sensor approach could provide fast and timely method for mineralogy tracking. Most of the
conventional spectroscopic methods in drill core analysis reveal their elemental information. This information can
be used to create models and predict the mineralogy, but it does suffer from the inability to detect the molecular
bonds and autonomously build precise predictions. Elemental information requires background information on the
sample mineralogy to be able to build the models and it can be disturbed by similar mineral consistencies. Raman
spectroscopy is a good support for data fusion with elemental spectroscopy such as XRF and LIBS, because it can
indicate the molecular bonds instead of elements [1]. However, the conventional Raman spectrometers have limited
amount of minerals they can distinguish, because the fluorescence of the sample can lower the detection sensitivity
or sometimes overflow the signal entirely. This problem is common especially in minerals. A solution for reducing the
elevated signal background comes in the form of time-gated Raman spectroscopy [2]. This technology uses short laser
pulses to excite the sample and measures the scattered radiation in short time-window before the slower timescale
phenomena, fluorescence, has time to manifest. In many cases, this can eliminate the fluorescence background
and bring on to view spectral features that are commonly hidden or weak using traditional Raman techniques. In
this study, timegated PicoRaman M2 spectrometer, utilizing 532 nm pulsed excitation laser, is integrated to a multi-
sensor Ancorelog drill core logger solution as a mineral identification tool. The measurement utilises an auto-focus
probe developed at VTT with ability to set the system in focus autonomously during measurements. In addition, a
mineral identification software is developed for automated identification of minerals using Raman spectroscopy. The
solution utilizes reference spectra from the RRUFF online database of minerals [3]. A classification model correlates
the measured spectra with reference spectra, which are selected using expert knowledge of the sample mineralogy.
The correlation utilizes a dot product between the spectra as a metric.
| Original language | English |
|---|---|
| Title of host publication | 11th International Conference on Advanced Vibrational Spectroscopy |
| Subtitle of host publication | Abstract Book |
| Pages | 198 |
| Number of pages | 1 |
| Publication status | Published - 23 Aug 2021 |
| MoE publication type | Not Eligible |
| Event | 11th International Conference on Advanced Vibrational Spectroscopy, ICAVS 11 - Online Duration: 23 Aug 2021 → 26 Aug 2021 |
Conference
| Conference | 11th International Conference on Advanced Vibrational Spectroscopy, ICAVS 11 |
|---|---|
| Abbreviated title | ICAVS 11 |
| Period | 23/08/21 → 26/08/21 |
Keywords
- Time-gated Raman spectroscopy
- Drill core analysis
- Ancorelog
- Raman mineralogy
- multi-sensor analysis