Product Note R35 - 08/16

BRAVO –
The must have handheld
Raman solution for
Pharmaceutical Industry

Advantages
Calcium Carbonate
Patented fluorescence mitigation Corn Starch
Croscarmellose Sodium
Large spectral range
Lactose
Laser class 1M – safe operation Microcrystalline Cellulose
Raman Intensity

Talc
Intuitive and guided touchscreen operation
Highest wavelength accuracy and precision
Enhanced material distinction
Compliant to latest Pharma regulations

Process and cost optimizations challenge our daily business. 3000 2500 2000 1500 1000 500
Raman Shift cm-1
By introducing handheld Raman spectrometers that enable to
probe materials like active pharmaceutical ingredients or exci-
Figure 1: BRAVO SSE TM spectra of typical Pharmaceutical excipients.
pients directly through transparent packaging at any location
new opportunities emerged.
With BRAVO the dedicated handheld Raman solution for the Duo LASERTM Excitation – Large Spectral Range
Pharmaceutical industry combining outstanding performance BRAVO allows recording a large spectral range from 300 cm -1
and ease of use Bruker started a new era in material control. to 3200 cm -1 Raman shift and hence provides even informa-
tion from the CH-stretching range (Figure 1).
SSETM – Patented Fluorescence Mitigation
Sequentially Shifted Excitation allows measuring more mate- Laser class 1M – Safe Operation
rials compared to conventional handheld Raman spectrome- As a laser class 1M product operation of BRAVO does not
ters by mitigation of fluorescence while keeping the perfor- require any laser safety equipment, restricted area or laser
mance at least as high as for benchtop systems (Figure 1). safety officer.
 Home screen with status Fully configurable can retrieve all Clear result presentaon and further funconalies as
informaon and material informaon with one scan. batch scan mode and advanced result presentaon
overview of workflows available.

Figure 2: Verification workflow

Intuitive and Guided Touch screen Operation HQI approach with BRAVO already results in a successful
Likewise the operation of a smartphone the user is guided distinction which can be significantly increased further by
through a clearly laid out user interface designed for the combination with SSE TM.
needs of material inspection. BRAVO provides highest stan-
dards and sophisticated workflows for efficient operation IntelliTipTM - Automated Tip Recognition
(Figure 2). When acquiring library spectra BRAVO stores tip informati-
on automatically and advises during the verification process
Material, Method and Library Setup which tip to be used. IntelliTipTM hence always ensures that
BRAVO makes building libraries quick and easy by an intui- the proper tip according to the material to be measured
tive workflow in combination with Bruker’s powerful OPUS is applied in order to receive best spectrum quality at any
software suite. Due to the high precision optics of BRAVO time.
best results in the market using an HQI approach are achie-
ved enhancing the distinction of even very similar materials. Validation
In Table 1 results from the verification of celluloses are
BRAVO and all of its software are fully compliant to the
presented comparing an HQI approach and an HQI approach
latest Pharmaceutical regulations like 21 CFR Part 11, EP
in combination with SSE TM. It can be clearly seen that an
2.2.48 and USP 1120. Due to the robust and precise optics
BRAVO’s wavelength precision and accuracy by far exceed
the recently specified regulatory requirements for handheld
Sample evaluated Correlation with SSETM Raman spectrometers and outperforms those of other
towards MCC (no SSETM) handheld Raman devices. In particular it can be noted that
microcrystalline cellulose 1 0.99 BRAVO’s performance is comparable to high performance
(MCC) benchtop Raman spectrometers and matches even the
Hydroxy propyl methyl 0.93 0.58 more stringent regulatory requirements of this instrument
cellulose (HPMC) class.
Hydroxy propyl cellulose 0.86 0.87 Reference standards for comprehensive system tests of
(HPC) BRAVO are according to ASTM 1840 as specified by the
Sodium carboxy methyl 0.97 0.66 USP and EP regulations and of course a comprehensive
cellulose (SCMC) validation manual including certificates and step by step
instructions for IQ, OQ and PQ procedures is available.
Table 1: Spectra of diffrent cellulose materials tested against a
method for the verification for MCC. The calculated correlation is
displayed with and without the use of the SSE method.

www.bruker.com/optics Bruker Optics

Bruker Optics is continually improving its products and reserves the right to change specifications without notice.
© 2016 Bruker Optics BOPT - 4000983-01
Product Note R34 - 05/15
BRAVO –
Advanced Data Acquisition
and Evaluation in Handheld
Raman Spectroscopy
The main benefit of handheld spectrometers is that the
instrument itself can be taken to the sample and thus can be
employed for various applications with high degree of flexi-
bility. For routine applications like materials identification and
verification, it is highly desirable to have as many functions as
possible automated. A good example is the quality control
of incoming goods using handheld Raman spectrometers,
which allows analyzing materials directly through packaging
[1].
There are also always many advanced applications where
the instrument requirements will be user or application
dependent. Previously most small handheld Raman systems
lacked the flexibility to allow the user to optimize the data
collection for the desired analysis. The handheld Raman
spectrometer BRAVO features an advanced operation mode
via remote control utilizing Bruker’s comprehensive spec-
troscopy software OPUS. The remote control is accomplis-
hed via Wi-Fi or an Ethernet connection using a docking sta-
tion. Consequently, in any case where a mobile or handheld
instrument needs to be run untethered to electrical power
supply, and has to require flexible data acquisition BRAVO is
a very powerful solution.
With BRAVO there is the direct option to initiate measure-
ments from the OPUS software and to subsequently run
user defined evaluation processes. In addition, it is possible
to set parameters such as integration time and number of
co-added measurements manually. The settings can be
stored in experiment files which can be assigned to a user
or specific application.
What possibilities are coming up?
BRAVO measures Raman spectra in a quality typically only
obtainable from powerful benchtop spectrometers [2]. This
means that acquired spectra can be used for any kind of
evaluation such as quantification or complex identification
methods. At this stage, not only the mathematically pro-
cessed data used for on board verifications based on the
Sequentially Shifted Excitation (SSETM) method is available
but also the raw data resulting from each laser excitation is
accessible, which is often preferred for scientific applica-
tions [3].
The spectroscopy software suite OPUS combines a wide
range of functionality for analytical and research applications
with an intuitive design and highest degree of flexibility.
Furthermore, additional software packages are available to
meet the needs of specific applications [4]. The following
examples provide insight into the many possibilities based
on the remote control option and comprehensive OPUS
spectroscopy software.
OPUS SEARCH
The OPUS SEARCH package is an advanced library
searching tool capable of mixture analysis and represents a
typical approach to identify the spectra of unknown materi- -0.5
als. A typical example is given in Figure 1. Libraries can be
setup by the user and many comprehensive databases from
common suppliers are available. Typical applications include
0.0
forensic investigations, detection of hazardous materials by
first responders, materials characterization in the industrial
PC3
environment, and analysis of objects in the field of art and
cultural heritage [5]. 0.5

OPUS IDENT
The OPUS IDENT package is an excellent comprehensive
tool for reliable identification of raw materials, intermedi-
-0.5 -0.25
ates and finished products offering the highest flexibility for 0.0
0.25
method setup. The combination of user defined data pre- 0.0
0.5
PC4
PC2
treatment, selection of various methods for data evaluation 0.5 0.75
1.0
and control of many more parameters make it the ultimate
easy-to-use program for dedicated quality control solutions.
Fig. 2: 3D score plot of an IDENT analysis of similar oil samples
Figure 2 shows the example of the differentiation of very measured with the BRAVO. The colored spheres represent the areas
similar oil samples using principle component analysis. in which the identification was successful, respectively. Very similar
materials (yellow and blue spheres) can be addressed in sub libraries
OPUS QUANT or assigned to one class.
OPUS QUANT is a state-of-the-art software package for
setting up and validating quantification models utilizing
advanced PLS based methods. Analogous to the setup of an 80 anhydrite content
cross validation
IDENT method, the user is guided step-by-step in generating 70
the quantification models. In Figure 3, an example of a cali-
Prediction (%)

60
bration is given to determine the relative anhydrite content 50
in gypsum. 40
30
20
10

 


  


$ % "&'
 #

10 20 30 40 50 60 70 80
True (%)

Fig. 3: Calibration model for the relative anhydrite content in gypsum

based on the evaluation of the symmetric SO 4 vibration.

References
[1] F. Fromm, New Possibilities in Raw Material ID
Using Handheld Raman Spectroscopy, Pharmaceutical
 Business Review, October 2015.
[2] Bruker Product Note T30 03/16, Accuracy is crucial:
       The starting point for a robust transfer of methods.
 ! "
# [3] Bruker Product Note T29 12/15, Efficient mitigation of
fluorescence in Raman spectroscopy using SSE TM.
Figure 1: Example of the analysis of an unknown liquid (red spec-
[4] https://www.bruker.com/opus
trum). The mixture analysis based on the analysis of multiple libraries
reveals the presence of two materials, polyethylene glycol (PEG) [5] Conti et al., Portable Sequentially Shifted Excitation
and the synthetic cannabinoid JWH-018. The calculated composition Raman spectroscopy as an innovative tool for in situ chemical
spectrum (dashed line) matches the query spectrum. interrogation of painted surfaces, Analyst (2016).

www.bruker.com/optics Bruker Optics

Bruker Optics is continually improving its products and reserves the right to change specifications without notice.
© 2015 Bruker Optics BOPT - 4000809 - 01
Product Note T30 03/16
BRAVO - Accuracy is crucial:
The starting point for a
robust transfer of methods
Handheld devices nowadays make Raman spectroscopy
available for various routine applications operated with a
high degree of automation. Some decades ago this could
not be imagined as for example it often was a challenge
to ensure and achieve the required Raman shift accuracy
during operation. With new arising developments, i.e. very
stable diode lasers, Raman spectroscopy is getting more
and more robust and user-friendly.
An important issue for handheld Raman spectrometers
is the capability of a robust transfer of libraries between
different instruments of the same type. In other words it is
required that the same results are obtained with a common
set of library data for various instruments. Prerequisite is a
thorough calibration which guarantees a high accuracy and
sets every spectrometer to a well-defined state. Of course,
the second requirement is the capability of a spectrometer
maintaining its performance during operation which is
related to a high precision. Especially this issue is of high
importance as handheld instruments are not operated under
laboratory conditions but instead being exposed to manifold
varying environmental conditions. Thus, a high accuracy
combined with a high precision form the basis for a reliable
in field operation using library data being generated in a
centralized manner on multiple spectrometers.
With BRAVO special care is taken to make the highest
standards only common for benchtop instrumentation
available in handheld Raman spectroscopy. X-axis
calibration is based on measurements of the emission of
a neon lamp and Raman shift standards as defined by the
ASTM [1]. Multiple Raman lines are evaluated to ensure a
comprehensive calibration valid across the entire spectral
range. To demonstrate the high level of accuracy achieved
table 1 compares peak positions of cyclohexane to ASTM
literature values based on measurements performed at
multiple BRAVO spectrometers.
Notably, all Raman lines match very closely their literature
values and the individual deviations are well below
1 cm -1 which is exceptional for handheld spectrometers.
To emphasize this results it needs to be mentioned that
cyclohexane was not used for spectrometer calibration.
For the Pharmaceutical industry the chapters USP 1120
by the United States and EP 2.2.48 by the European
Pharmacopeia define the minimum system requirements
and guidelines for Raman instrumentation being operated in
validated environments [2, 3].For example the EP 2.2.48 as
well defines acceptable tolerances for the x-axis accuracy
which are given in table 1 for benchtop and handheld
instrumentation. It is obvious that the BRAVO spectrometer
 Table 1:

Cyclohexane - Cyclohexane – BRAVO Deviation to ASTM Allowed deviations according to

ASTM values [1] (average +/- std. dev.) EP 2.2.48 (benchtop/handheld)
801.3 801.2 +/- 0.20 0.08 1.5 / 2.5
1028.3 1027.9 +/- 0.20 0.35 1.0 / 2.0
1266.4 1266.3 +/- 0.19 0.074 1.0 / 2.0
1444.4 1443.9 +/- 0.22 0.61 1.0 / 2.5
2852.9 2853.6 +/- 0.29 -0.66 2.0 / 3.0

Table 1: Comparison of Raman peak positions (unit: cm -1) of cyclohexane to ASTM literature values as determined by BRAVO. The positions
are shown as average values and standard deviation based on single measurements at 20 different BRAVO spectrometers, respectively.

is capable to be operated in line to requirements for OPUS the overall performance of the spectrometer can
benchtop instrumentation and easily exceeds the softened be easily monitored running comprehensive system tests
values for handheld instrumentation. according to current Pharmaceutical regulations.

A common way to compare spectral data for library

matching is to calculate the correlation coefficient between

1.00
the measured spectrum and reference data. The correlation
coefficient calculates to a number ranging from 1 to -1
0.95
representing a qualitative measure of identity in which 1
typical threshold
Correlation coefficient

represents a perfect match. In reality this parameter is

not only influenced by the similarity of the reference and
0.90

analyzed material but depends as well on other parameters

such as background signals, i.e. fluorescence. In general it is
aimed to add as much as possible significance to this value
0.85

in means of selectivity towards other materials, for example

an adequate data pretreatment is recommended [4].
0.80

Regarding this aspect BRAVO’s patented SSETM mitigates

fluorescence signals in an active manner [5].
calcite
(400 cm-1 to 1800 cm-1)
0.75

A further influence on the correlation is given by the

accuracy of a spectrometer which is illustrated in figure 1.
Here the correlation coefficient of two Raman spectra of
calcite is calculated in the range of 400 cm -1 to 1800 cm -1
as function of a constant deviation in x-axis accuracy across
the entire spectral range. It is evident that for a deviation
of 2 cm -1 the typically applied threshold of 0.95 is barely
reached.
To ensure the required x-axis accuracy for BRAVO not only
at an instrument test but continuously during operation a
neon spectrum is generated with every Raman spectrum
measured using an inbuilt neon lamp. Of course, the
same care is taken at y-axis calibration stage using a NIST
SRM relative intensity correction standard for Raman
spectroscopy. Finally, with the spectroscopy software
0 1 2 3 4 5
Deviation (cm-1)
Figure 1: Correlation coefficient of two spectra of calcite in the range
of 400 cm -1 to 1800 cm -1 as function of their relative deviation in x-
axis accuracy (constant shift across the spectral range).
References
[1] ASTM E1840-96(2014), Standard Guide for Raman Shift
Standards for Spectrometer Calibration.
[2] <1120> Raman Spectroscopy, United States Pharmacopeia.
[3] 2.2.48. Raman Spectroscopy, EUROPEAN PHARMACOPEIA
8.7.
[4] J. Kauffman et al., „Spectral Processing for Raman Library
Searching“, Amer. Pharm. Rev. 14, (2011).
[5] Bruker Product Note T29 12/15, “Efficient mitigation of
fluorescence in Raman spectroscopy using SSE TM”.

www.bruker.com/optics Bruker Optics

Bruker Optics is continually improving its products and reserves the right to change specifications without notice.
© 2016 Bruker Optics BOPT - 4000899-01
Product Note T29 12/15
BRAVO
Efficient mitigation of
fluorescence in Raman
spectroscopy using SSETM
For many years, Raman spectroscopy has been utilized
as an important tool for materials characterization in the
analytical laboratory. With new developments in laser
technology and spectrometer design, handheld Raman
spectrometers can now be designed and built for practical
everyday use in a vast array of applications [1]. The BRAVO
by Bruker takes handheld Raman spectroscopy to a new
level, overcoming the limitations of previously available
systems. These previous limitations included i.e., limited
wavelength accuracy [2], non-safe laser usage, and
fluorescence interference. The wavelength accuracy of
the BRAVO is significantly better than competitive devices
resulting in the highest data consistency. The BRAVO
is also a Class 1M laser safe device. Lastly, the BRAVO
incorporates new technology called Sequentially Shifted
Excitation (SSE TM ) to mitigate fluorescence.
The general task of a handheld instrument is obvious – it
is a spectrometer which can be employed anywhere with
the highest degree of flexibility, where the sample in
question does not need to be transferred to a laboratory for
reliable and rapid identification. A requirement that goes
hand-in-hand with the automation of measurements and
data analysis in handheld devices is to achieve a maximum
robustness with respect to factors affecting the reliability
of results. This starts with minimizing the risk for operating
errors with intelligent software that provides an intuitive
guided workflow for acquiring and processing the data.
One of the most challenging aspects of Raman analysis
is overcoming fluorescence interference. Fluorescence
illumination is a highly efficient process that frequently
yields a strong signal that can overwhelm the desired
Raman spectrum. Additionally, a fluorescence background
can limit reproducibility, which is required for conducting
analyses in the materials identification and quantification
process. The patented Sequentially Shifted Excitation
(SSE TM ) method used by BRAVO detects and removes
the fluorescence signal leaving only the desired Raman
spectrum [3,4].
How does SSETM work?
In order to extract the Raman signal from an acquired
spectrum, the BRAVO software must recognize and
differentiate the Raman spectrum from other interfering
signals (fluorescence) automatically. To make this
possible the wavelength of the lasers used for excitation
are sequentially shifted during the measurement. As a
consequence, the position of the Raman signal on the
detector is deliberately changed and the signals at constant
wavelength, such as fluorescence, remain at the same
position. The SSE TM algorithm extracts the “moving” signals
which relate directly to the Raman signature (see Fig. 1)
yielding the desired spectrum.
 talc in PE bag
Raman intensity (arb. units)
raw data
at SSE TM absorpon
SSETM
processed PE signals
400 600 800 1000 1200
Raman shi (cm-1)
Figure 1: Raman spectra of talc in a PE bag at sequentially shifted ex-
citation (raw data at SSE TM ) and after fluorescence mitigation (SSE TM
processed). Raman lines of talc indicated by dashed lines are moving
as the laser wavelength is changed. The SSE TM processed spectrum
only contains the “moving” signals extracted from the three spectra.
What are the benefits of SSETM?
One of the main advantages of SSE TM is that after the data
is processed, the resulting spectrum is free of fluorescence
and is most suitable for unambiguous identification. Without
SSE TM, a data pretreatment such as baseline correction
would be necessary in order to attempt to remove the
interfering fluorescence signal. The possible information
loss because of a data pretreatment could yield invalid
results [5].
The SSE TM method is most sensitive in retaining the
intrinsic bandshape and positions of bands related to
the material of interest. This allows not only the quick
and accurate identification of unknown compounds, but
also the characterization of contaminants and other trace
compounds. Figure 2 shows an example where sodium
alginate acquired with 785 nm excitation has an intense
fluorescence background dominating the spectrum of
interest (red spectrum). After application of the SSE TM
technique with the BRAVO (blue spectrum), the resulting
spectrum can be ready analyzed and the compound reliably
identified.
Some handheld Raman devices employ 1064 nm
lasers, because fluorescence is generally less common
with NIR excitation. However, the Raman scattering
efficiency decreases significantly as longer wavelength
lasers are used. The resulting lower signal needs to be
www.bruker.com/optics
sodium alginate
Raman intensity (arb. units)
785 nm
excitaon
BRAVO
1400 400 600 800 1000 1200 1400
Raman shi (cm-1)
Figure 2: Comparison of Raman spectra of sodium alginate acquired
using a conventional dispersive Raman instrument at 785 nm excita-
tion (red) and BRAVO (blue).
compensated for with increased measuring time and/or
higher laser power. Additionally, many samples undergo
transient heating with 1064 nm excitation. SSE TM handles
fluorescence while maintaining high sensitivity and under
the safest conditions: The BRAVO is a class 1M system in
all modes of operation.
Consequently, whenever fluorescence or background
signals are present, SSE TM offers significant benefits for
unambiguous material identification of the broadest range of
materials.
References
[1] S. Assi, „Identification of counterfeit drugs using dual laser
handheld Raman“, Eur. Pharm. Rev. 20, 20 (2015)
[2] Product Note T30: BRAVO - Accuracy is crucial: The starting
point for a robust transfer of methods
[3] J. B. Cooper et al., “Sequentially Shifted Excitation Raman
Spectroscopy: Novel Algorithm and Instrumentation for
Fluorescence-Free Raman Spectroscopy in Spectral Space”,
Appl. Spectrosc. 67, 973 (2013)
[4] J. B. Cooper et al., “Spatially compressed dual-wavelength
excitation Raman spectrometer”, Appl. Opt. 53, 3333 (2014)
[5] J. Kauffmann et al., “Spectral Preprocessing for Raman
Library Searching”, Amer. Pharm. Rev. 14, (2011)
Bruker Optics
Bruker Optics is continually improving its products and reserves the right to change specifications without notice.
© 2016 Bruker Optics BOPT - 4000884-02